University of Groningen Offspring of subfertile couples ... · Cover design: Rinze Schols, drawings by Rinze Schols & Pamela Schendelaar, 2014 . Thesis lay-out: Pamela Schendelaar

University of Groningen

Offspring of subfertile couples: neurodevelopmental outcome at preschool ageSchendelaar, Pamela

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2015

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Schendelaar, P. (2015). Offspring of subfertile couples: neurodevelopmental outcome at preschool age.[S.l.]: [S.n.].

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 09-10-2020

https://www.rug.nl/research/portal/en/publications/offspring-of-subfertile-couples-neurodevelopmental-outcome-at-preschool-age(9af0d6f2-2f1f-4a32-a660-5769cb1b7974).html

https://www.rug.nl/research/portal/en/publications/offspring-of-subfertile-couples-neurodevelopmental-outcome-at-preschool-age(9af0d6f2-2f1f-4a32-a660-5769cb1b7974).html

Offspring of subfertile couples: neurodevelopmental outcome

at preschool age

Pamela Schendelaar

The studies in this thesis have been carried out with financial support from the University of Groningen, the University Medical Center Groningen (grant number 754510), the Junior Scientific Masterclass, the Postgraduate School of Behavioural and Cognitive Neurosciences, the Netherlands Organization for Health Research and Development (ZonMw, grant number 945-03-013) and the Cornelia Foundation. The printing of this thesis was financially supported by the University of Groningen, the University Medical Center Groningen, Ferring B.V., Goodlife Pharma, Noord Negentig and Origio Benelux B.V.

Their support is gratefully acknowledged.

Offspring of subfertile couples: neurodevelopmental outcome at preschool age © Copyright 2015, P. Schendelaar, the Netherlands. All rights reserved. No part of this thesis may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system, without prior written permission of the author or, when appropriate, of the publishers of the publications. ISBN: 978-90-367-7560-1 ISBN electronic version: 978-90-367-7559-5 Cover: A toddler’s hand reaches for the pine cone which is held by its mother. Reaching and grasping small objects is part of the neurological examination to evaluate a child’s fine manipulative abilities. The pine cone was a familiar fertility symbol in ancient times. In this drawing, handing over a pine cone like if it is a toy, symbolizes the shift from the wonder and mystery of reproduction and pregnancy towards something that is nowadays more understandable, tangible and verifiable. Cover design: Rinze Schols, drawings by Rinze Schols & Pamela Schendelaar, 2014 Thesis lay-out: Pamela Schendelaar Printed by: Studio Michał Sławiński, thesisprint.eu

Offspring of subfertile couples: neurodevelopmental outcome

at preschool age

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen

op gezag van de rector magnificus prof. dr. E. Sterken

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

maandag 16 februari 2015 om 14.30 uur

door

Pamela Schendelaar

geboren op 2 december 1987 te Amersfoort

Promotores Prof. dr. M. Hadders-Algra Prof. dr. M.J. Heineman Copromotor Dr. K.J. Middelburg Beoordelingscommissie Prof. dr. O.F. Brouwer Prof. dr. S.A. Scherjon Prof. dr. F. van der Veen

Paranimfen Jorien Seggers Anne van Zoonen

If you can dream it, you can do it

- Walt Disney

voor papa & mama

voor oma voor Henna

CONTENTS CHAPTER 1 General introduction ................................................................................................ 11 Part I: The Groningen ART cohort study

CHAPTER 2 .................................................................................................................................. 33 Movement variation in infants born following IVF/ICSI with and without ovarian hyperstimulation CHAPTER 3 .................................................................................................................................. 45 The Groningen ART cohort study: the effects of ovarian hyperstimulation and the IVF laboratory procedures on neurological condition at 2 years CHAPTER 4 .................................................................................................................................. 61 Increased time to pregnancy is associated with suboptimal neurological condition of 2-year-olds CHAPTER 5 .................................................................................................................................. 67 Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not CHAPTER 6 ................................................................................................................................. 87 Subfertility factors rather than assisted conception factors affect cognitive and behavioural development of 4-year-old singletons

Part II: The PGS study

CHAPTER 7 ................................................................................................................................. 101 The effect of preimplantation genetic screening on neurological, cognitive and behavioural development in 4-year-old children: follow-up of a RCT

CHAPTER 8 General discussion ................................................................................................. 119 CHAPTER 9 Summary ................................................................................................................ 139 CHAPTER 10 Nederlandse samenvatting (Summary in Dutch) ................................................... 145 APPENDICES References................................................................................................................................. 153 Dankwoord (Acknowledgements) ................................................................................................ 171 Over de auteur (About the author) ............................................................................................. 177 List of publications .................................................................................................................... 179

11

CHAPTER 1 General introduction

12

Women’s role in society has changed considerably over the last decades. In modern society women are educated equivalent to men and more often pursue a successful career. As a consequence of the increasing economic independence, women postpone having children. The latter is also attributable to the development and availability of better contraceptive methods. Nearly three decades ago two-thirds of Dutch women gave birth below the age of 30 years; nowadays, only one-third of women does.1 The mean maternal age at first birth has increased by about four years between 1970 and 2011 in the United States (from 21.4 years to 25.6 years) and is even higher in developed European countries, with the Netherlands as one of the outliers with an average maternal age of 29.4 years.1,2 The shift towards a higher maternal age at child conception has its impact on the population’s fertility. Fertility clearly declines after the age of thirty, with a more progressive decline in the late thirties.3-5 Moreover, older women who conceive are at greater risk of pregnancy complications.6,7

While the women’s emancipation has taken its effort during the second half of the twentieth century, medical science has gained momentum as well. In 1976 the gynaecologist Patrick Steptoe and the physiologist Robert Edwards took the first steps in the development of assisted conception by achieving the first pregnancy after transfer and implantation of a human embryo which was fertilized in vitro.8 In 1978 they gained international attention by reporting the first birth of a baby after in vitro fertilization (IVF), named Louise Brown.9 Since that date, up to 5% of newborns in Europe.10 and 1% of newborns in the United States11 are born following assisted reproductive techniques (ART). In the Netherlands, up to 3% of children are born following ART.12

With the introduction of ART, the responsibility follows to carefully monitor and warrant its safety. The health and development of ART offspring is of general interest, since these children nowadays represent a substantial part of the world’s population.

In this thesis the neurodevelopmental outcome of children born following ART is evaluated up until 4 years of age by exploring the influence of specific factors involved in assisted conception, such as ovarian hyperstimulation, the in vitro laboratory procedures and subfertility-related aspects. Prior to the content of the thesis, the applied fertility treatment techniques are introduced briefly and a general and concise overview of the literature available on child health and developmental outcome is given.

Assisted reproductive techniques Over the years, many assisted reproductive techniques have been developed, with a wide variation in medical indications and invasiveness of the procedures. IVF is considered to be the mainstay of technologies in the field of assisted conception. The techniques that were applied in the studies of this thesis will be outlined briefly. No specific attention will be paid to fertility treatments such as intra-uterine insemination (IUI) and ovulation induction (OI).

1

General introduction

13

Conventional controlled ovarian hyperstimulation IVF In vitro fertilization is the process of fertilizing oocytes outside the human body in the laboratory after which one or more embryos are transferred into the uterine cavity. The conventional procedure, controlled ovarian hyperstimulation IVF (COH-IVF), consists of several phases. It starts with ovarian hyperstimulation, in which the ovaries are stimulated by means of daily injections of exogenous follicle stimulating hormone (FSH) to induce growth of multiple follicles. These injections are combined with the supplementation of either gonadotropin-releasing hormone (GnRH)-agonists or GnRH-antagonists to prevent the endogenous luteinizing hormone (LH) surge and thus, premature ovulation. Next, follicle growth is carefully monitored by ultrasound. In case of a follicle diameter of approximately 18 to 20 mm, either human chorionic gonadotrophin (hCG) or LH is administered to facilitate follicle maturation. About 34 to 36 hours later, one or more oocytes are retrieved from the ovarian follicles by means of transvaginal ultrasound-guided aspiration. Subsequently, the retrieved oocytes are mixed with spermatozoa in a small volume of culture medium, or a single spermatozoon is directly injected into the cytoplasm of the oocyte. The latter technique is known as intracytoplasmic sperm injection (ICSI). ICSI is mostly applied in case of male subfertility.13 After fertilization, the resulting zygotes are maintained in culture for two to three days to develop into a cleavage stage embryo. Approximately two to five days after fertilization, one or two morphologically good embryos are transferred into the uterine cavity (single or double embryo transfer, respectively SET and DET). Prior to the transfer, progesterone or hCG may be administered to support the endometrium. The embryos of good quality that are not transferred to the uterus can be cryopreserved for future transfer in subsequent cycles.14

Modified natural cycle IVF A disadvantage of COH-IVF is the increased chance of having a multiple pregnancy, which in turn is associated with obstetric and perinatal adversities.15,16 Another disadvantage of COH-IVF is the chance of developing ovarian hyperstimulation syndrome (OHSS), a rare but potentially life-threatening complication of ovarian hyperstimulation. In OHSS a combination of ovarian enlargement occurs, due to multiple ovarian cysts and an acute fluid shift out of the intravascular space.17,18 To reduce the risk of such complications, an alternative for conventional IVF was developed: modified natural cycle IVF (MNC-IVF). In contrast to COH-IVF, in MNC-IVF the one follicle that naturally develops as dominant is used for conception. Again, follicle growth is carefully monitored by means of ultrasound. However, in the late follicular phase of the natural cycle, when the leading follicle has reached a diameter of approximately 14 mm, GnRH antagonists and FSH are administered, in most cases resulting in the maturation of a single follicle.19 Benefits of MNC-IVF are the maintenance of the natural selection of the follicle and the minimal use of hormonal medication. A drawback, however, is the low chance of an ongoing pregnancy per IVF treatment cycle of approximately 8%.20

1

14

Preimplantation Genetic Screening Preimplantation genetic screening (PGS) was introduced to enhance efficiency of assisted conception. The success rate of IVF is relatively low, with approximately 50% chance of an ongoing pregnancy after three IVF treatment cycles. In PGS, embryonic cells are screened for aneuploidy and only embryos with euploid cells are selected for transfer to the uterus. Aneuploid cells contain an abnormal number of chromosomes: the cell either contains additional chromosomes or is missing chromosomes, a phenomenon that occurs during cell division when the chromosomes do not separate properly. In order to screen for aneuploidy, one or two blastomeres (i.e. cells produced by division in a fertilized ovum) are aspirated from a cleavage stage embryo. To this end, an opening is created in the embryos zona pellucida by enzymatic digestion or by laser. After removal of one or two blastomeres fluorescent in situ hybridization (FISH) is applied to identify the copy numbers of several sets of chromosomes in the cell.

In theory, PGS results in improved implantation, increased ongoing pregnancies, and live-birth rates, as it is assumed that about half of all embryos obtained through IVF are aneuploid.21 Unfortunately, PGS was less effective in improving ongoing pregnancy rates than expected. In fact, PGS even significantly reduced the live-birth rate for women of advanced maternal age.22,23 As a consequence, PGS as described in the present thesis is no longer applied. The European Society of Human Reproduction and Embryology PGD Consortium recommends applying PGS only in the context of properly constructed trials.24

Embryo biopsy is currently still applied in the form of preimplantation genetic diagnosis (PGD). In PGD, embryos of couples with a significant risk of a genetic disorder are analysed for inheritance. The application of PGD is steadily increasing, underlining the importance of evaluating the effects of embryo biopsy.

Assisted reproduction – the potential interference with child development in general The idea of potential influences of assisted conception on child development can be traced to the ‘developmental origin of health and disease paradigm’, formerly known as the 'Barker hypothesis', which states that the early development of an organism serves as a template for its future health, or, in other words: ‘the early environment shapes an individual’s health later in life’.25-27 During critical phases of an organism’s early development, functional and structural organic changes take place depending on the environmental circumstances in which the organism lives. Such physiologic and metabolic alterations may predispose organisms to increased susceptibility to disease and developmental disturbances later in life. A number of animal studies have indeed demonstrated that the periconceptional environment is decisive for the individual’s later quality of life.28-30

ART may interfere with early human development during several steps in human reproduction. First, ovarian hyperstimulation causes a down-regulation of the pituitary

1


15

function and stimulates the ovaries to produce an increased number of follicles. As a result of the growth of multiple follicles, natural selection of the dominant follicle may be bypassed. Other points of concern are the fertilization and maturation of the oocyte and the embryo in vitro and the manipulation of the oocyte and embryo both mechanically and chemically by means of the use of cultured media.31,32 Several animal studies demonstrated that in vitro culture impairs embryonic development, pregnancy establishment and outcome.33 Cultured media have been linked to alterations in gene imprinting in mice and abnormal fetal outcome, including high and low birthweight.33-35 The scarce literature on humans shows conflicting results. A study by Dumoulin et al. showed that culture conditions can affect birthweight of newborns.32 The authors compared perinatal outcome of children born following IVF in which either one of two widely used, commercially available culture media were used (Vitrolife AB media, Göteborg, Sweden; Cook media, Brisbane, Australia). They found significantly higher pregnancy rates and clinical pregnancy rates in the Vitrolife culture group and a significant higher birthweight in the Vitrolife culture group (mean birthweight: Vitrolife: 3453; Cook: 3208) than in the control group exposed to the Cook medium. Later on, the authors found similar results in a study using frozen embryos and the same culture media (adjusted mean difference birthweight: 122 g, P = 0.03). However, they were not able to find an effect of culture media on gestational age.36 Other groups were not able to find associations between cultured media types and birthweight and gestational age.37,38 In addition, Lin et al. did not find significant associations between birthweight and gestational age and the type of culture media (G5TM, Global and Quinn’s advantage media) in 1201 singletons and 445 sets of twins.39

Another source of potential developmental interference of IVF may be the endometrium, which is subjected to an altered endocrinological environment due to hormonal stimulation.40 Normal preimplantation development and establishment and maintenance of a pregnancy rely on mutual communication between the young embryo and adjacent uterine epithelial cells that coordinate embryonic development with consecutive changes in uterine function. Several animal studies demonstrated that ovarian hyperstimulation is associated with alterations in hormonal levels and disturbances in the endometrium due to, among other things, a non-optimal expression of endometrial growth factors and reduced endometrial receptivity.41,42

An increasing body of evidence suggests that ART is associated with epigenetic changes, i.e. modification of chromosomes that do not alter the nucleotide base sequences but alter the expression of genes.33,43 Disruptions in genomic imprinting is one of epigenetic processes that seems to be associated with ART.43,44 Genomic imprinting is an epigenetic process of the differential expression of genetic material depending on whether it was inherited from the mother or father, with the silent allele based on the parent-of-origin. Theoretically, ART could interfere with genomic imprinting through ovarian hyperstimulation and follicle aspiration. Germ cells have the ability to erase the imprinting marks during their development in embryonic life. After resetting, differentiating germ cells re-establish de novo imprinting marks according to their sex. The latter occurs for both

1

16

sexes in late fetal stages and continues after birth. In oocytes, the process is not completed until just before each ovulation, whereas spermatozoa complete the process of re-establishing paternal imprinting marks earlier in the gametogenesis. Both mouse and human oocytes from stimulated cycles appear to have altered methylation status of several imprinted genes compared to oocytes from non-stimulated cycles.45 Several epidemiological studies indicate an increase in the incidence of imprinting disorders in children born following ART compared to naturally conceived (NC) children, such as Beckwith-Wiedemann syndrome and Angelman’s syndrome.43 Moreover, not only ART aspects seem to interfere with genomic imprinting. Some studies reported associations between a history of subfertility and a higher prevalence of imprinting disorders.46,47 Although interesting, imprinting disorders are beyond the scope of this thesis on child neurodevelopment.

The following paragraphs focus on the health and development of children born to subfertile couples, with or without the application of ART. First, the general health and development will be described, from birth until puberty, in terms of congenital abnormalities, cardiometabolic and cardiovascular health, and growth and pubertal development. Thereafter, a shift is made towards the main focus of this thesis: neurodevelopment after ART and subfertility. The literature available on child neurological development will be outlined in terms of neuromotor, cognitive and behavioural development.

Assisted reproduction – child health and developmental outcome in general Over time, concerns have been raised about the safety of ART with regard to child outcome. ART is known to be associated with multiple gestations, which is in turn known to be associated with an increased risk for obstetric and perinatal adversities, in particular, preterm delivery and low birthweight. The latter are associated with increased perinatal mortality and short-term and long-term morbidity.15,16 In addition, multiple gestation may also be associated with the early loss and absorption of one twin, known as the vanishing twin syndrome.48 A little more than 10% of IVF singletons originates from a twin gestation early in pregnancy,49 which may contribute to adverse developmental outcome after IVF.50,51

Despite the fact that ART is inextricably linked to the increased number of twins, it is well established that the numerous adverse perinatal outcomes, which are associated with ART, cannot be ascribed only to the effects of multiple gestation. Also singletons born following ART are two times more often born preterm and more often have low birthweight compared to NC peers.52-56 Moreover, ART singletons are three times more often very preterm and have very low birthweight, even when factors such as maternal age at conception and parity were taken into account.52-56 Moreover, ART singletons are more often born following Caesarean section and more often referred to the neonatal intensive care unit (NICU). The risk of perinatal mortality among ART singletons is almost twofold compared to NC peers.52-56 Evidence is accumulating that not only the ART aspects interfere with child outcome, but inherent factors of the treated couples, such as increased maternal

1


17

age7,57,58 and a history of subfertility,59-62 may also contribute to unfavourable developmental outcome. Subfertility is known to be associated with obstetrical problems such as the need of Caesarean section, pregnancy-induced hypertension or preeclampsia, placenta praevia and placental abruption and increased perinatal adversities such as preterm birth, low birthweight, and perinatal deaths.61-63

Congenital abnormalities Another point of concern regarding health outcome after assisted conception is the fact that children born following ART are at increased risk for congenital abnormalities compared to their NC peers. Hansen et al. pooled the estimates derived from the meta-analyses of six systematic reviews on birth defects in ART children compared to NC peers.64 The pooled odds ratios (ORs) show a 30 to 70% increase in birth defects (OR: 1.3 to 1.7), with slightly lower pooled ORs for studies examining singletons and multiples combined (OR: 1.3 to 1.4). A recent population-wide cohort study by Davies et al. demonstrated an increased risk of congenital abnormalities in children born following IVF with ICSI and not in those born following IVF per se.65 The increased risk persisted after adjustment for maternal age and several other risk factors.

Whether the increased risk of congenital abnormalities can be attributed to ART aspects or the underlying parental subfertility is unclear; the literature shows conflicting results. Davies et al. found a higher prevalence of congenital abnormalities among children born following ART compared to their NC siblings, suggesting that ART-related aspects may be the decisive factor.65 Zhu et al. on the other hand, found that severe subfertility was associated with a higher prevalence of congenital abnormalities.66

Cardiometabolic and cardiovascular health Recent evidence suggests that, to some extent, ART children are presumably more prone to develop cardiometabolic diseases due to the increased risk of perinatal adversities,52-56 which in turn are risk factors for type 2 diabetes, hypertension, and cardiovascular disease.25-27 However, the increased risk may also be related to epigenetic modifications as a result of the ART procedure per se or the underlying parental subfertility.67,68 Several animal studies have shown that in vitro embryo culture contributes to significantly heavier offspring than spontaneous conception and that ART may be associated with impaired glucose metabolism and elevated blood pressure levels in offspring, adversities that were found in human ART children as well.68,69 It has been suggested that ART children display a level of vascular dysfunction similar to that seen in children of mothers with preeclampsia. For example, Scherrer et al. found an indication for generalized vascular dysfunction in healthy children conceived by ART.70 With the study of Scherrer et al. included, only a handful of studies have reported worse cardiometabolic health in children born following ART.70-75 Several studies reported subtle, yet significantly elevated blood pressure levels in ART children compared to NC peers born to fertile71,75 or subfertile couples,73 independent of early life factors and parental characteristics. One study even reported elevated blood

1

18

pressure levels and increased aortic intima-media thickness after ART in utero.76 However, Belva et al. reassessed their ART children at older age, resulting in the disappearance of the previously found difference in blood pressure levels, presumably because of the influence of puberty.72 Nevertheless, the authors did find higher amounts of total body fat, particularly in ICSI girls compared to NC peers.72 Also Miles et al. could not detect a disadvantageous effect of ART and reported an even favourable lipid profile in pre-pubertal ART children.59,77 On the contrary, Ceelen et al. reported, besides the aforementioned elevated blood pressure levels, thicker skinfolds and higher fasting glucose levels in ART children. However, no differences were found in BMI or fasting insulin concentrations and insulin resistance measures compared to NC peers born to subfertile parents.73,74 Given the scarcity of literature available on cardiometabolic health among ART children, further research needs to be done, especially given the long-term consequences of cardiometabolic and cardiovascular risks.

Growth and pubertal development Several cohort studies from Western-European countries have studied long-term growth and physical development of children and have not found clear differences between ART and NC children, both in singletons and multiples. The weight, height, and head circumference of IVF and ICSI children at various ages, varying from 1 to 8 years, have been reported to be normal.78-82 Considering the fact that ART children are at risk for perinatal adversities such as low birthweight and prematurity, the disappearance of impaired growth with increasing age implies catch-up growth. Indeed, several of the former reported studies reported the occurrence of catch-up growth in ART children.78-80 However, one study observed sustained growth retardation in both IVF singletons and multiples compared to natural conceived controls at the age of 3 years, presumably due to insufficient catch-up growth after a period of poor perinatal outcome.79 Catch-up growth seems positive, but it may have consequences for later life. In SGA born children it has been linked to early onset obesity with the subsequent emergence of (components of) the metabolic syndrome.83,84 The higher prevalence of perinatal adversities in ART children increases their chance of experiencing catch-up growth. Consequently, the ART population may indirectly be more prone to develop health problems later in life.

As much is known about birthweight and growth of ART children, so little is known about the later pubertal development. The majority of people who were born following ART have not reached adulthood until now and only a small group of offspring has reached the age of 30 and beyond. According to some studies, ART does not appear to affect the onset of puberty.75,85,86 However, there are some suggestions that ART may interfere at a different level in puberty. For instance, Belva et al reported delayed breast development in 14-year-old ICSI girls, however, the onset of the pubarche and menarche was similar to that of NC peers.87 Ceelen et al. reported an advanced bone age in pubertal IVF girls, but not for IVF boys, compared to NC peers.85 Moreover, the authors found higher serum levels of dehydroepiandrosterone sulphate (DHEAS, a steroid hormone, that functions as a

1


19

metabolic intermediate in the biosynthesis of the androgen and oestrogen) and LH in IVF girls. The clinical relevance of these higher values need to be elucidated, but according to the authors the elevated DHEAS and LH levels could be due to perinatal growth restriction and the high variation in pulsatile LH secretion in the girls, respectively.85 A study by Mau Kai et al. reported disadvantageous effects of ART for boys, in terms of lower serum testosterone concentrations and a higher LH/testosterone ratio in ICSI boys compared to NC boys.88 No differences were found between IVF boys and NC boys. According to the authors, the ICSI boys may have inherited the subtle impairment of Leydig cell from their fathers.88

Assisted reproduction – neurodevelopmental outcome of the child Neurodevelopment is a broad term and generally used as an umbrella term for the growth and development of the brain and central nervous system. Terms such as impaired neurodevelopment or neurodevelopmental disorders or disabilities are used to refer to brain dysfunction that affects motor function, emotion and behaviour, learning ability, self-control and memory.

A thoroughly performed systematic review by Middelburg et al. in 2008 indicated that ART has not been associated with adverse neurodevelopmental outcome during the first postnatal years.89 It also revealed that in 2008, surprisingly little was known on long-term outcome due to the relative paucity of long-term follow-up studies of good methodological quality. Early development is only a moderate predictor of long-term development, meaning that reassuring findings on short-term neurodevelopmental outcome do not preclude an effect of ART or subfertility on long-term neurological development, as it takes time for some disorders to emerge.90

After 2008, various studies on long-term neurological development after assisted conception, beyond the age of 18 months, have been performed. However, consensus on whether or not ART has an impact on long-term neurodevelopmental outcome is still lacking. The results of the studies vary due to methodological issues such as diversity in study outcome and instruments used to assess outcome, the lack of a suitable proper control group, high rates of attrition, the variety in age at which the subjects were studied and the type of study subjects that were studied, such as singletons, multiples or both. An overview of the current literature on long-term neurodevelopmental outcome, i.e. outcome beyond the age of 18 months, is presented in the following paragraphs. Neuromotor development There are some concerns about the potential risk of cerebral palsy (CP) in children born following ART. CP is an umbrella term for a group of permanent clinical disorders that are characterized by motor and postural dysfunction. These conditions, which range in severity, are due to non-progressive lesions or abnormalities of the developing brain resulting from a variety of causes, of which the appearance may change over time as the brain matures.91,92

A systematic review and meta-analysis on the association between CP and fertility treatment were performed by Hvidtjorn et al. on 41 published studies.93 The authors found

1

20

an increased risk for CP in IVF children compared to NC peers, which may be attributable to prematurity, multiplicity, and the vanishing twin phenomenon in early pregnancy. Soon after, Hvidtjorn et al. performed a population based follow-up study including nearly 600,000 children to assess the risk of CP in IVF children in comparison to NC children.94 The authors found that children born after IVF had an increased risk of a CP diagnosis (hazard ratio (HR) [95% confidence interval (CI)]: 2.34 [1.81-3.01]) compared to NC children, a significant finding that disappeared after adjustment for multiplicity and gestational age. Hereafter, a publication by Zhu et al., followed, using register-based data. The authors found that children born after ART had an increased risk of CP, even after adjustment for prematurity and multiplicity (HR [95% CI]: 2.30 [1.12–4.73]).95 The authors did not find a significant association between time to pregnancy (TTP) and the risk of CP in children conceived spontaneously.95

Various studies reported on the less severe but more common neurodevelopmental disorders in ART children. The studies, however, vary considerably in the nature of ART, assessment method, age at follow-up and in study outcome. Some studies reported worse neuromotor outcome in ART children,96-98 whereas others concluded that outcome of ART children was similar to that of NC children.78,99-106 Few studies addressed neurological development beyond the age of 18 months. Zhu et al. found a delay for most milestones, especially motor milestones, in 18-month-old ICSI children.97 On the other hand, Ludwig et al. reported that neuromotor skills of 5.5-year-old ICSI singletons were similar to those of NC peers.105 Similar results were found by Leunens et al. in 8 and 10-year-olds.107,108 Both latter studies, however, suffered from a high attrition rate. Knoester et al., who assessed minor neurological dysfunction (MND, see page 29) at 5 to 8 years, reported that the neurological condition of children born after IVF was similar to that of NC children.104 However, ICSI children had a higher prevalence of MND than the NC children. Finally, Levy-Shiff et al. performed a general paediatric neurological examination in 9 to 10-year-old children and found no differences between IVF/ICSI children and NC children.100

Thus, the effect of ART on long-term neuromotor is still unclear. The picture is complicated by a potential effect of subfertility. As mentioned before, suggestions have been made for the effect of factors inherent to treated couples on neurodevelopment in the offspring, such as increased maternal age7,57,58 and a history of subfertility.59-62 For instance, Zhu et al. found a modest increased risk for developmental coordination disorder (DCD) in 7-year-old children born to subfertile couples. The majority of studies does not properly distinguish between potential ART-related effects and subfertility aspects on child neurodevelopmental outcome due to the design of the study groups. They compare ART outcomes with those of fertile populations rather than to the results of a subfertile population conceiving spontaneously and, consequently, overestimate the potential effect of the fertility treatment. A more proper control group consists of naturally conceived children born to subfertile couples, as it resembles an ART population more in terms of parity and parental age.

1


21

Cognitive development The majority of long-term follow-up studies on cognitive outcome in ART children reported similar cognitive functioning compared to NC peers99-102,105,108-111 or even better cognitive functioning.107,112 The better cognitive functioning found in the latter two studies in 8-year-old ICSI children and 8 to 17-year-old IVF children in comparison to NC children was presumably the result of more favourable parental characteristics in the ART population than in the spontaneously conceived controls.107,108,112 In contrast, Knoester et al. reported that cognitive function of 5 to 8-year-old singleton ICSI children was worse than that of NC children matched for gender, socioeconomic status, the presence of prematurity and maternal age at time of conception and date of birth.113 Goldbeck et al. reported lower IQ scores in 5 to 10-year-old ICSI singletons (mean IQ = 94.1, standard deviation [SD] = 13.8) compared to IVF peers (mean IQ = 102.0, SD = 9.1; P = 0.005). Moreover, they reported that 23.5% of ICSI children, but only 2.9% IVF children had at least borderline delayed cognitive development.114 In a study on language assessment in 5-year-old singletons and twins by Gucuyener et al., it was demonstrated that twins performed worse on the Stanford-Binet Intelligence Scale Form and the Peabody Picture Vocabulary test, especially when the twins were born after IVF.115 Two studies paid attention on the effect of a history of subfertility on the child’s cognitive outcome. Zhu et al. reported that a longer TTP may be associated with a delay in achieving certain milestones, in particular cognitive and language development.97 It appeared that maternal age and parity were the most consistent factors associated with infertility or fertility treatment.97 Carson et al. performed a prospective population based cohort study and investigated whether pregnancy planning, TTP and ART treatment influenced cognitive development in 3 and 5-year-old children.116 In contrast to the findings of Zhu et al., the authors found that these factors did not adversely affect the child’s cognitive development at age 3 or 5 and that the slight differences found in the unadjusted statistical analyses – findings in favour of the ART children – were entirely attributable to socioeconomic differences among the participants.

Behavioural development As already mentioned, ART is associated with perinatal adversities such as low birthweight and preterm delivery.52-56 These perinatal adversities, in turn, are, together with other ART-related risk factors such as advanced maternal age, unwanted childlessness, and obesity, associated with neurodevelopmental disorders,117 such as attention deficit hyperactive disorder (ADHD) and autism spectrum disorders (ASD).118,119 Therefore, the main focus of studies on behavioural development in ART children is on ADHD and ASD. The review by Middelburg et al. concluded that ART is not associated with adverse behavioural development during the first postnatal years.89 But again, early development is not a good predictor for the development of the aforementioned minor developmental problems.90 A review by Wagenaar et al. reported no differences in the prevalence of behavioural and socio-emotional problems in ART and NC children up to the age of 8.120 Hereafter, several studies reported additional reassuring findings on long-term behavioural outcome, with

1

22

similar behavioural outcome in ART children compared to NC children.86,102,117,121-125 However, some unfavourable findings were reported as well in IVF children, varying from more vocalization and higher energy levels,99 withdrawn and depressive behaviour126 and behavioural problems, particularly in boys.100 A pilot study by Zachor and Itzcak on the association between assisted conception and the risk of ASD in 8 months to 18-year-old children demonstrated that ART appears to be a significant independent risk factor for ASD.127 Beydoun et al. evaluated the quality of life and susceptibility for chronic disease development in 21-year-old ART children by means of self-administered questionnaires.86 The authors reported a slightly increased risk of depression (15.6% compared to the expected 12.7% lifetime prevalence up to age 25) and a significantly increased prevalence of female binge drinking in the ART group compared to a control group (55% versus 37%). However, this may not be fully representative for the ART population as less than one-third of participants completed the questionnaire.

The overview indicates that we are not well informed about the long-term consequences of assisted conception and subfertility-related aspects on a child’s neurodevelopment. We only have sufficient information on short-term neurodevelopmental outcome. However, as indicated before, early development is only a moderate predictor of long-term development, as children may grow into neurodevelopmental deficits at older age.90 In order to understand the latter concept, I will now briefly summarize brain development.

Brain development The human nervous system already starts to develop when the embryo is merely 2 weeks old. Around 3 to 4 weeks of embryonic life a hollow cylindrical structure, the neural tube, is formed, originating from the ectodermic neural plate, the source of the majority of neurons and glia cells in the mature human.128-130 The tube sinks under the surface of the skin, the forward end enlarges and differentiates into the forebrain, midbrain, and spinal cord. The fluid-filled cavity within the neural tube becomes the central canal of the spinal cord and the four ventricles of the brain, with the fluid being the cerebrospinal fluid. Its interaction with the surrounding mesoderm gives rise to the dura, the cranium, and the vertebrae. A process of neuronal proliferation, migration, organization and myelination follows, events that span a period from 9 weeks to adult life.128-130 Brain mass develops by the production of new cells, i.e. cell proliferation, from particular regions of the expanding neural tube, mainly between 8 to 16 weeks. Some cells continue to divide and redivide as stem cells, whereas other cells differentiate into neurons and glia cells, the main cellular components of the brain. Mainly at 12 to 20 weeks, the newly generated neurons and glia cells migrate towards particular areas of the outer layer of the neural tube by following specific chemical paths of immunoglobulins and chemokines, eventually forming the cortex and subcortical nuclei. By 20 to 24 weeks, the human cerebral cortex essentially has its full complement of neurons. Hereafter, a delicate process of organizational events occur, mainly from 20 weeks to several years after birth. Once the neurons have reached their final destination in the outer

1


23

layer, they extend axons and dendrites and connect them to synapses. The resulting white matter represents the cortical connectivity. Dendritic development accelerates from the third trimester onwards to the end of the first postnatal year and on a lower level till about the age of 5.131 Like all neurodevelopmental processes, the progress of the dendritic differentiation depends on the establishment of afferent input and presumably synaptic activity. Synaptic formation differs considerably among brain regions. The synaptogenesis starts early during fetal life with a peak at 34 to 36 weeks and some decline after birth.128-130 Initially, synapses make unspecified functional connections, resulting in an overproduction of synapses. Some of the synaptic contacts are utilised in emerging neuronal circuits, whereas those that remain unspecified are eliminated by means of programmed cell death, i.e. apoptosis.128 Not only synapses are eliminated. Selective elimination of other neuronal processes co-occur. The phase of such regressive events is critical to eliminate abnormal projections and to match the number of incoming axons to the number of receiving cells.129,130 Even though apoptosis depends on endogenous programmed processes, animal experiments indicate that enriched behavioural experience during the period of programmed cell death results in an increased level of neural cells in adult life.131 Next to the synaptic reorganisation, myelination is another long-lasting human neurodevelopmental event, that starts in the second pregnancy trimester and continues throughout a major part of adult life.128-130 Myelination involves the producing of insulating fatty sheaths around axons by glia cells in order to accelerate transmission.

Thus, it becomes clear that the development of the human nervous system is a delicate, dynamic and ongoing process that already starts early in gestation and continues into adulthood, with different brain areas maturing at different times and rates.128,131 Between birth and adulthood the human brain increases in size from around 100 grams at birth to 1000 grams by the end of the first year, close to the adult weight of 1200 to 1500 grams.128-130 The resulting continuous synaptic reorganisation forms the basis of neural development and brain plasticity. Experience modifies development: within limits, the human brain has evolved the ability to remodel itself by reorganizing neural pathways in response to experience. The continuous functional and structural changes during pre- and postnatal life have important clinical consequences and implications for the prediction of functional outcome.90,128,131,132 The age-specific brain of the child determines the way in which neural (dys)function is expressed. Children can only display (minor) deviations in functions which belong to their age-specific abilities. In addition, specific neurodevelopmental events result in specific vulnerability for impaired neurodevelopmental outcome, depending on the maturation and development of the brain at the time of the event. Circumstances that would almost certainly lead to severe neurological dysfunction in an adult can have quite different consequences for children. Moreover, children that initially have normal neurological function at an early age may grow into a functional neurological deficit, as with increasing age the complexity of neural functions increases. A simple example of such a ‘growing in a deficit’ phenomenon may be dyslexia, a disorder that only comes to light when children start learning reading and writing. Neurodevelopmental

1

24

changes may result in the opposite as well, meaning that early dysfunction may disappear at later age. The latter is seen during puberty, which is related to neurological improvement, presumably due hormonal influences.133 Adolescence is a sensitive period for steroid-dependent brain organization.133 The elevated levels of gonadal steroid hormones trigger the (re)organisation of neural circuits and increase the white matter integrity.134

Aim of the thesis The general aim of this thesis is to evaluate the effects of assisted conception and subfertility-related aspects on neurodevelopmental outcome in preschool-aged children, ranging from 4 months to 4 years of age. Data of two parallel running projects, the Groningen ART cohort study and the PGS study, are used. Both projects will be outlined below. The specific research questions addressed in this thesis are: • Do ART-related aspects such as ovarian hyperstimulation, the in vitro laboratory

procedures, or a combination of both affect neurodevelopmental outcome o in terms of movement variation in children of 4, 10 and 18 months of age? o in terms of neuromotor function in children of 2 and 4 years of age? o in terms of cognition and behaviour in children of 4 years of age?

• Do subfertility-related aspects such as the presence of a history of subfertility, the underlying cause of subfertility and the severity of subfertility affect neurodevelopmental outcome in terms of neuromotor function, cognition and behaviour in children of 2 and 4 years of age?

• Does PGS in addition to conventional IVF or ICSI affect neurodevelopmental outcome in terms of neuromotor function, cognition and behaviour in children of 4 years of age?

The Groningen ART cohort study The Groningen ART cohort study is a prospective assessor-blinded longitudinal follow-up study that primarily aims to evaluate the effects of ovarian hyperstimulation, the in vitro procedure, and the combination of both on neurodevelopmental outcome in children born following ART, including both IVF and ICSI. Pregnant subfertile couples with a term date between March 2005 and December 2006 were recruited at the Department of Reproductive Medicine of the University Medical Center Groningen (UMCG).135 All couples who achieved a singleton pregnancy following IVF or ICSI were invited to participate in the study. This resulted in a group of children born following controlled ovarian hyperstimulation IVF or ICSI (COH-IVF) and a group of children born following IVF or ICSI in a modified natural cycle (MNC-IVF), in which no ovarian hyperstimulation was applied and the follicle that developed naturally to dominance was used for assisted conception (Figure I and II).136,137 All oocytes were cultured in human tubal fluid (HTF) medium (Lonza, Verviers, Belgium) and supplemented with 10% plasma solution (Sanquin, Amsterdam, The Netherlands). Couples that were treated with cryopreserved or donated oocytes or embryos to achieve pregnancy were excluded. A third group was formed by naturally conceived children born to subfertile parents (Sub-NC). These couples had tried to conceive

1


25

for at least one year and eventually conceived naturally while on the waiting list for fertility evaluation or treatment (Figure I and II). Potential differences in neurodevelopmental outcome of COH-IVF and MNC-IVF children may largely be attributed to ovarian hyperstimulation, whereas potential differences in MNC-IVF and Sub-NC children may largely be attributed to the in vitro procedure (Figure I).

The secondary aim of the Groningen ART cohort study is the evaluation of the effect of three aspects of subfertility on neurodevelopmental outcome in children born following ART: the presence of subfertility, the underlying cause of subfertility and the duration of subfertility.

In order to evaluate the effect of the presence of a history of subfertility, a reference group was retrospectively recruited at the assessment ages of 2 and 4 years, consisting of children born to fertile parents who conceived naturally (Figure I and II). The reference groups were recruited at six child welfare centres in and around Groningen between February and October 2009 for the 2-year assessment and December 2009 and February 2012 for the 4-year assessment. All parents of the children who visited the child welfare clinics for routine general health care were invited to participate. Children of parents who had tried to achieve a pregnancy for more than one year or who achieved a pregnancy by any form of assisted conception were excluded.

The underlying cause of subfertility and the duration of subfertility were studied in the three ART study groups. The causes of subfertility were classified as 'tuba pathology' in case of abnormalities of the fallopian tubes, as ‘male factor’ in case of male infertility, as ‘other causes’ in case of endometriosis, cervical factor or hormonal cause and as ‘unknown cause’ in case of lack of a specific cause for subfertility. The duration of subfertility was expressed in terms of time to pregnancy (TTP) as a proxy for the severity of subfertility. The notion that TTP can be used as a proxy for the severity of subfertility is based on models that have shown that TTP is an important predictor of the chance of pregnancy in subfertile couples.138,139 TTP was defined as the interval between the start of timed unprotected intercourse or a previous pregnancy and conception, recorded in years and months and finally converted into decimal years.

So far, the Groningen ART cohort study revealed that ART has not been associated with adverse neurodevelopmental outcome up until the age of 18 months,135,140 i.e. at the ages of 2 weeks, and 3, 4, 10 and 18 months.135,140 At the assessment age of 3 months it was possible to compare developmental outcome of the Groningen ART cohort children with that of a reference group. This group consisted of a representative sample of the general Dutch population, as the infants had been assessed as a part of a general health check-up provided for all infants. The data showed that neurological condition of the Groningen ART cohort children at 3 months of age was less favourable than that of 3-month-olds of the general population, suggesting that neurological sub-optimality at 3 months in children born after ART may rather be related to factors associated with subfertility than to ART techniques per se.135,140

1

26

1

FIGURE I. The Groningen ART cohort study. The effects of ovarian hyperstimulation, the IVF laboratory procedure, the combination of both and a history of subfertility are studied by four different comparisons. The groups controlled ovarian hyperstimulation IVF (COH-IVF), modified natural cycle IVF (MNC-IVF) and natural conception in subfertile couples (Sub-NC) were pooled to form the subfertile group.


27

FIGURE II. Flow chart of the Groningen ART cohort.

The PGS study The PGS study is a multicentre double-blind randomized clinical trial performed in the UMCG and Academic Medical Center (AMC), originally elaborated to evaluate the effectiveness of PGS by comparing ongoing-pregnancy rates after IVF with and without PGS in women of advanced maternal age (‘PGS for aneuploidies in IVF’, ZonMw project number 945-03-013).22 Inclusion criteria for participating women were a conception age ranging from 35 to 41 years, having no previously failed IVF-cycles and having no objections against a possible double embryo transfer. Randomization of women was performed centrally, using a computer program, with minimization for age (35-37 or 38-41 years) and reproductive technology (IVF or ICSI), and with stratification according to study centre prior to the start of the IVF procedures. Group status was revealed to participating parents after 12 weeks of gestation. Linked to this study, a follow-up study of children born after PGS was started in which all children born to women included in the double-blind randomized PGS-trial were examined

1

28

for the physical and mental health during childhood (Figure III). Prior to inclusion and randomization in the original PGS trial, couples were informed about the follow-up evaluation as part of the PGS trial. For practical reasons, children born after treatment in the UMCG were invited to participate in an extensive follow-up program similar to that of the Groningen ART cohort at the ages of 2 weeks, 3, 4, 10, 18 months, 2 and 4 years, whereas children born after treatment in the AMC were only invited for assessments at the age of 2 and 4 years.

So far, the PGS study revealed that neurodevelopmental outcome up until 2 years of age in children born following IVF with PGS was largely similar to that of children born following IVF without PGS.141,142 However, application of the detailed neurological optimality score revealed that children born after IVF with PGS had a somewhat less optimal neurological condition at 2 years than children born after IVF without PGS. This may suggest a less favourable neurological development in children born following PGS.

Neurodevelopmental assessments and outcome parameters In children, it is important to perform an age-specific assessment, since the young brain is continuously subjected to developmental changes. The following assessments were applied in the studies in this thesis at different ages.

Infant Motor Profile (IMP) The IMP is a video-based instrument that evaluates neuromotor condition in terms of the quality of spontaneous motor behaviour in infants aged 3 to 18 months.143 Its 80 items are organized in five domains, three traditional neuromotor domains (symmetry, fluency and performance) and two novel domains, variation and variability, that are based on the Neuronal Group Selection Theory (NGST).144,145 The first novel domain, variation, denotes the size of a child’s movement repertoire. Evidence is accumulating that movement variation is a better predictor of developmental outcome than the traditional neurological examination.146 Reduced variation is associated with early lesions of the periventricular white matter.145,147-149 and, more generally, may reflect the integrity of cortical connectivity.150 Moreover, reduced variation during infancy may be associated with neurodevelopmental disorders later in life145,151 and reduced intelligence at school age.152 The second novel domain, variability, reflects the ability to select adaptive motor strategies out of the repertoire. Variability in motor behaviour emerges during the first 18 postnatal months and is often deviant in children with a developmental disorder, regardless of the absence or presence of overt brain injury.145 The limited ability to select the best strategy for a specific situation is primarily due to deficiencies in the processing of sensory information associated with the child’s motor actions.145,147

In infants aged 3 to 6 months the total IMP score consists of the mean of four domain scores, i.e. all domains except the variability domain. In infants older than 6 months the total IMP score consists of the mean of all five domain scores, i.e. it includes also the variability domain. Prior and up to 6 months the infant’s limited variability precludes

1


29

inclusion of the variability domain into the total score.153 Total IMP scores and domain scores were expressed in percentages, varying from 0 to 100%. The reliability of the IMP is good and the construct validity is satisfactory.143,154

The neurological examination according to Hempel At 2 and 4 years, the neurological examination according to Hempel was applied. This is a standardized, precise and age-specific tool to assess minor neurological dysfunction (MND) at preschool age by means of five domains of functions: fine motor function, gross motor function, posture and muscle tone, reflexes and visuomotor function.155 Each of the domains can be scored as typical or deviant.

Children are classified as being neurologically normal, having simple MND, having complex MND, or being neurologically abnormal. Neurologically normal implies the absence of neurological dysfunction; it implies the absence of deviant domains or the isolated presence of dysfunction in the domain of reflexes. Simple MND indicates the presence of one deviant domain (except the domain of reflexes) and is regarded as a non-optimal but normal form of brain function.90 Complex MND indicates the presence of more than one domain of dysfunction. It represents the clinically relevant form of MND, as it is associated with prematurity and other perinatal adversities, and learning and behavioural disorders at later age.90,156-159 Neurologically abnormal implies the presence of a distinct neurological syndrome such as CP.

The outcome of the Hempel assessment was also expressed in a neurological optimality score (NOS). The NOS consists of 58 items (range 0 – 58) for 2-year-olds and 56 items (range 0 – 56) for 4-year-olds, for which an optimal condition is defined. The total score results from the sum of the items which fulfil the criteria for optimality. Higher scores represent better performance. Note that the range for optimal behaviour is narrower than for normal behaviour.160 The application of the optimality concept turns the NOS into a sensitive tool to evaluate neurological integrity. The fluency score (2-year-olds: range 0 – 13, 4-year-olds: 0 – 15) is a subscore of the NOS that evaluates the fluency of motor behaviour. The fluency score is a sensitive measure to detect subtle changes in neuromotor development, as minor dysfunction of the nervous system already results in a reduction of fluency of motion.161,162 The inter-rater reliability of the Hempel assessment is satisfactory (κ = 0.62 – 1.00 [mean 0.93]) and its construct validity is good.155,163-165

The Kaufman Assessment Battery for Children, second edition (K-ABC-II) The K-ABC-II is an individually administered standardized clinical instrument that measures cognitive and processing abilities in children and adolescents in the age of 3 to 18 years.166 Cognitive and processing abilities are expressed in a total intelligence quotient (IQ) score (i.e. Fluid-Crystallized Index) and four IQ scale scores: 1) a sequential processing IQ, reflecting short-term memory, 2) a simultaneous processing IQ, reflecting spatial aptitude, 3) a learning ability IQ, reflecting long-term memory capacity and 4) a knowledge IQ, reflecting general knowledge. Raw test scores are normalized into global scores with a

1

30

mean of 100 (SD: 15). The K-ABC-II has been used for children with different social backgrounds or ethnic differences without critical effects on test-scores. Reliability and validity of the K-ABC-II are good.166 The original American norms were applied as Dutch norms are lacking.

The Child Behavior CheckList (CBCL) The CBCL is a parental questionnaire which is used to identify emotional and behavioural problems in children aged 1.5 to 5 years.167 The questions on the CBCL are classified into the following problem scales: emotionally reactive, anxious/depressed, somatic complaints, withdrawn, sleep problems, attention problems and aggressive behaviour. The first four scales together form the internalizing scale; the latter two scales form the externalizing scale. The sum of all questions results in the total problem scale. Raw test scores are normalized into T-scores with a mean of 50 (SD: 10). Higher T-scores represent more problematic behaviour: T-scores < 60 are in the normal range of behaviour, T-scores between 60 and 63 represent borderline behaviour, and T-scores > 63 are in the clinically abnormal range of behaviour. In the papers of this thesis, the T-scores were used in the analyses. The reliability and validity of the CBCL are good.167 In this thesis, the validated Dutch version of the CBCL was used.168

Outline of the thesis The evaluation of the potential effects of ART and subfertility-related aspects on neurodevelopmental outcome in preschool-aged children is divided in two sections, related to the aforementioned research questions.

Part I: The Groningen ART cohort study Chapter 2 describes the neuromotor development in terms of movement variation in the children of the Groningen ART cohort at the age of 4, 10 and 18 months. Chapter 3 describes the neurological condition of the children of the Groningen ART cohort at the age of 2 years in terms of neurological optimality (fluency score, NOS) and MND. Potential effects of the presence and the underlying cause of subfertility on child neurological condition are evaluated. Chapter 4 describes the effect of the duration of subfertility (in terms of TTP, a proxy for the severity of subfertility) on the neurological condition of the children of the Groningen ART cohort at 2 years, again, in terms of neurological optimality and MND. Chapter 5 describes the effect of both ART-related aspects such as ovarian hyperstimulation and the in vitro laboratory procedures, and subfertility-related aspects, such as the presence, the underlying cause, and the severity of subfertility on neurological condition of the children of the Groningen ART cohort at the age of 4 years, in terms of neurological optimality and complex MND. Chapter 6 describes an explorative approach on the effect of both ART-related aspects such as ovarian hyperstimulation and the in vitro laboratory procedures, and subfertility-related

1


31

aspects, such as the presence, the underlying cause and the severity of subfertility on cognitive and behavioural development of the children of the Groningen ART cohort at the age of 4 years. Causal inference search algorithms and structural equation modelling were applied as statistical tools in order to unravel underlying causal mechanisms and to distinguish in intermediate and confounding effects.

Part II: The Groningen PGS study Chapter 7 describes the effect of preimplantation genetic screening in addition to IVF on neurodevelopmental outcome, in terms of neuromotor, cognitive and behavioural development in children of the PGS study at the age of 4 years.

Chapter 8 consists of a general discussion of the findings, the final conclusions, the strengths and limitations of the studies in this thesis, and future perspectives. Chapter 9 and 10 summarize the content of this thesis in English and Dutch respectively.

1

32

Movement variation in infants born following IVF/ICSI with and without ovarian hyperstimulation

33

CHAPTER 2 Movement variation in infants born following IVF/ICSI with and without ovarian hyperstimulation

Pamela Schendelaar

Kirsten R. Heineman

Maas Jan Heineman

Marjolein Jongbloed-Pereboom

Sacha la Bastide-van Gemert

Karin J. Middelburg

Edwin R. van den Heuvel

Mijna Hadders-Algra

Early Human Development 2013 ; 89 : 507 – 513

34

ABSTRACT

Background The effect of in vitro fertilization (IVF) on neurodevelopment is not fully understood. Probably, IVF does not affect traditional measures of neurodevelopment in infancy. Recently, an instrument, the Infant Motor Profile (IMP), was developed that evaluates the quality of motor behaviour. It includes the evaluation of movement variation (i.e. movement repertoire size), a parameter reflecting the integrity of cortical connectivity. Aim To evaluate the effect of ovarian hyperstimulation and the in vitro procedure on movement variation during infancy. Study design Prospective cohort study. Subjects Singletons born following IVF or intracytoplasmic sperm injection (ICSI) with conventional controlled ovarian hyperstimulation (COH-IVF, n=68), in a modified natural cycle (MNC-IVF, n=57) and natural conception born to subfertile couples (Sub-NC, n=90). Outcome measures Children were assessed with the IMP at 4, 10 and 18 months, resulting in a total IMP score and five domain scores: variation, variability, symmetry, fluency and performance. Primary outcome was the domain score variation. Results A significant effect of study group was observed for the variation score up until 18 months of age (P = 0.039). COH-IVF children had a significantly lower mean variation score than MNC-IVF children (mean difference [95% confidence interval] -1.010 [-1.766 to -0.254]). Mean variation scores of COH-IVF and Sub-NC children were similar; the same held true for the comparison between MNC-IVF and Sub-NC. Total IMP scores and other domain scores of the three groups were similar. Conclusion The present study did not demonstrate a clear effect of ovarian hyperstimulation and the in vitro procedure on movement variation throughout infancy.

2


35

INTRODUCTION

Assisted reproductive techniques (ART) such as in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) are known to be associated with perinatal complication such as preterm birth and low birthweight.52-55 Nevertheless, so far neurodevelopmental outcome during the first postnatal years of singletons born after ART appears to be similar to that of singletons born after natural conception.89 This finding is reassuring, but recently evidence is accumulating that ART may have an effect on the child’s later health.169,170 Hence, continuing critical evaluation of the effect of ART on neurodevelopmental outcome is warranted.

Hypothetically, several components of assisted conception may interfere with child development, including brain development, such as the impact ovarian hyperstimulation on oocytes and endometrium171 and the in vitro procedure in terms of embryo culture and manipulation.31 Other associations with assisted conception that may contribute to an adverse developmental outcome are factors such as andrological factors, an increased maternal age7,57 and a history of subfertility.59-62

We recently developed a new tool to assess the infant’s neuromotor function, the Infant Motor Profile (IMP).143 The IMP is a video-based instrument that evaluates neuromotor condition in terms of the quality of spontaneous motor behaviour in infants aged 3 to 18 months.143,144 Its 80 items are organized in five domains, three traditional neuromotor domains (symmetry, fluency and performance) and two novel domains, variation and variability, that are based on the Neuronal Group Selection Theory (NGST).144,145 The first novel domain, variation, denotes the size of a child’s movement repertoire. Evidence is accumulating that movement variation is a better predictor of developmental outcome than the traditional neurological examination.146 Reduced variation is associated with early lesions of the periventricular white matter145,147-149 and – more generally – may reflect the integrity of cortical connectivity.150 Moreover, reduced variation during infancy may be associated with neurodevelopmental disorders in later life145,151 and reduced intelligence at school age.152 The second novel domain, variability, reflects the ability to select adaptive motor strategies out of the repertoire. Variability in motor behaviour emerges during the first 18 postnatal months and is often deviant in children with a developmental disorder, regardless the absence or presence of overt brain injury.145 The limited ability to select the best strategy for a specific situation is primarily due to deficiencies in the processing of sensory information associated with the child’s motor actions.145,147

To study potential effects of the different components of ART, we composed the Groningen ART cohort.135 This cohort consists of the following three groups: a group of children born following IVF or ICSI with conventional controlled ovarian hyperstimulation (COH-IVF) – the conventional form of IVF in which ovarian hyperstimulation induces growth of multiple follicles, a group of children born following IVF or ICSI in the modified natural cycle (MNC-IVF) – in which no ovarian hyperstimulation is applied and the one follicle that naturally develops to dominance is used for treatment,136,137 and a group of naturally

2

36

conceived children born to subfertile couples (Sub-NC). Potential differences between COH-IVF and MNC-IVF children are suggestive for an effect of ovarian hyperstimulation, whereas potential differences between MNC-IVF and Sub-NC children will reflect the effect of the in vitro procedure.

We previously reported that neurological condition expressed in terms of minor neurological dysfunction and neurological optimality scores did not differ between the three groups of the Groningen ART cohort up to and including the age of 2 years.135,140,172 Nevertheless, it remains possible that the various components of ART affect brain development.

The present study aims to assess neuromotor development of the children of the prospective Groningen ART cohort at 4, 10 and 18 months of age with the IMP. Primary outcome is the variation domain score at 4, 10 and 18 months. Potential differences between the three groups are most likely to occur in the variation domain as this domain differs most clearly from the neurological concepts evaluated previously with the traditional neurological examination in terms of minor neurological dysfunction and neurological optimality score. Additional outcome parameters are the quality of spontaneous motor behaviour expressed in the total IMP score and the other domain scores variability, fluency, symmetry and performance.

MATERIALS AND METHODS Participants Pregnant subfertile couples with a term date between March 2005 and December 2006 were recruited at the Department of Reproductive Medicine of the University Medical Center Groningen.135 All couples who achieved a singleton pregnancy following IVF or – for example in case of severe male subfertility – following ICSI were invited to participate, which resulted in children born following COH-IVF and MNC-IVF. Couples with a pregnancy after treatment with donated oocytes or embryos or cryopreservation were excluded. A third group resulted in children born naturally to subfertile couples (Sub-NC). These couples had tried to conceive for at least one year, and finally conceived naturally while waiting for fertility evaluation or treatment.

Prenatal, perinatal and demographic information was gathered by use of standardized charts during the first follow-up assessment, approximately 2 weeks post-term.135 When information was incomplete or complications during pregnancy or birth had occurred, midwifes and gynaecologists were asked for details. Detailed information on the causes and treatment of infertility was retrieved from medical records.

The Medical Ethical Commission of the University Medical Center Groningen approved the study design. Parents provided written informed consent for participation of their child in the study.

2


37

Motor assessment The quality of motor behaviour was assessed with the IMP at 4, 10 and 18 months.143 The IMP assessment is a video-based instrument that evaluates the quality of spontaneous motor behaviour in infants aged 3 to 18 months. It consists of 80 items that are scored in different positions and situations, i.e. in supine, prone, sitting, standing and walking, and during reaching and grasping. These scores result in five domain scores: variation (i.e. the size of movement repertoire), variability (i.e. the ability to select motor strategies), symmetry, fluency and performance. In infants aged 3 to 6 months the total IMP score consists of the mean of four domain scores, i.e. all domains except the variability domain. In infants older than 6 months the total IMP score consists of the mean of all five domain scores, i.e. it includes also the variability domain. Prior and up to 6 months the infant’s limited variability precludes inclusion of the variability domain into the total score.153 Total IMP scores and domain scores were expressed in percentages, varying from 0 to 100%. The reliability of the IMP is good and the construct validity is satisfactory.143,154

The videos were assessed by four trained observers (M.H.-A., K.R.H., M.J.-P., A.S.) Prior to study onset inter-observer reliability was tested. To this end, three assessors independently analysed 16 videos at 4 and at 10 months and two assessors independently analysed 21 videos at 18 months. The intraclass correlation coefficients for the various domains indicated satisfactory to good reliability (Table I). During the entire study all assessors were blind to prenatal and perinatal history, including the mode of conception.

Statistical analysis Mann-Whitney U-tests and Chi-square or Fisher’s exact tests were used to test differences in demographic characteristics between groups. The numerical outcome variables, i.e. the scores of the domains variation, variability, fluency, symmetry and performance and the total IMP score, observed at 4, 10, and 18 months of age for each child were analysed with a population-averaged or marginal mixed model under the assumption of normality. The three-dimensional unstructured covariance matrix for the outcome variable over age was estimated per study group (COH-IVF, MNC-IVF or Sub-NC) using restricted maximum

Primary outcomeVariation score 0.80 (0.79 - 0.91)

Additional outcomesTotal IMP score 0.95 (0.80 - 0.95)Variability score 0.79 (0.60 - 0.88)Symmetry score 0.99 (0.64 - 1.00)Fluency score 0.74 (0.60 - 0.81)Performance score 0.98 (0.80 - 0.99)

TABLE I. Inter-observer reliability of the total IMP and domain scores.

Intraclass correlation coefficient (range)

2

38

likelihood. The fixed effects in the marginal model incorporated age as a second order polynomial to address differences in individual ages for children at 4, 10, and 18 months. This quadratic relation was fitted per study group, implying an average age profile specific to study group. Subsequently, the overall effect of study group was adjusted for sex, gestational age, birthweight, vanishing twins, time to pregnancy, signs of fetal distress and pregnancy-induced hypertension.

The likelihood ratio test (LRT) was first applied to investigate the effect of study group on the covariance structure at 5% significance level. The same covariance structure was taken for the groups when the LRT did not demonstrate significance. Secondly, the LRT was applied to test if study group would have an effect on the mean level of the outcome variables by fitting the models with the selected covariance structure using maximum likelihood. The LRT was applied to test if a separate profile in age was indeed necessary for study group. Again a significance level of 5% was selected. When this LRT did not demonstrate significance on the average age profiles, the LRT was applied to test if the mean level of the outcome variable was affected by study group. In case the last test would be significant at 5% significance level, study group would have a consistent effect on the mean level of the outcome variable over the investigated age period when corrected for the other variables, i.e. an effect of the groups is present but unrelated to age. When study group had no average age specific profile, the mean difference between the groups were estimated and accompanied with an approximate 95% confidence interval using Satterthwaite173 to the calculation of the number of degrees of freedom.

All analyses were performed with the Statistical Analysis System (SAS) software, version 9.2 and the Predictive Analytics software (PASW) Statistics, version 18. In all analyses probability values of 5% or less were considered significant.

RESULTS

Participation and demographic characteristics Eighty-nine children born after COH-IVF, 79 MNC-IVF children and 143 Sub-NC children met the inclusion criteria during the prenatal period. Parents of 68 (76%), 57 (72%) and 90 (63%), respectively, agreed to participate in neurodevelopmental follow-up.135 Overall, obstetrical, neonatal and social characteristics of participants and non-participants were similar, except for maternal age, which was significantly lower for non-participating Sub-NC mothers (P = 0.030).135

Demographic and perinatal characteristics of participants are outlined in Table II. In general, the groups were similar. However, birthweight and gestational age were significantly higher and longer in Sub-NC children compared to COH-IVF children (birthweight: P = 0.020; gestational age: P = 0.020). Signs of fetal distress were observed more frequently in Sub-NC than in MNC-IVF children (P = 0.046). Sub-NC mothers had more often pregnancy-induced hypertension compared to MNC-IVF mothers (P = 0.043). Time to

2


39

pregnancy was shorter in the Sub-NC group compared with the COH-IVF group (P < 0.001) and MNC-IVF/ICSI group (P = 0.002). Eight COH-IVF children were survivors of a vanishing twin, whereas one MNC-IVF child was (P = 0.040) and no Sub-NC children were (P = 0.001). None of the children who had been admitted to the neonatal care unit had a serious brain lesion, e.g. cystic periventricular leukomalacia and intraventricular haemorrhage grades III or IV (data not presented).

Attrition at the assessment at 18 months was minimal: two COH-IVF, one MNC-IVF and four Sub-NC children were lost to follow-up. Five of these children were not assessed due to logistical reasons. One MNC-IVF girl died of a congenital heart disorder when she was 3 weeks old. One Sub-NC boy was assessed at 18 months, but his data were excluded from the analysis as his walking behaviour was clearly affected by a fibular aplasia. Eventually, video-recordings of 66 (97%) COH-IVF children, 56 (98%) MNC-IVF children, and 86 (96%) Sub-NC children were analysed at 4, 10 and 18 months (overall attrition rate 3%).

Variation at 4, 10 and 18 months At 4 months of age, the mean variation scores of the COH-IVF, MNC-IVF and Sub-NC children were 91.6, 94.8 and 94.5 respectively (for details see Table III). At 10 months, the mean variation scores were 98.2, 98.3 and 98.4 for COH-IVF, MNC-IVF and Sub-NC children, respectively and those at 18 months 96.3, 98.1 and 97.0, respectively (for details see Table III).

The mixed model indicated an overall effect of the study groups on the variation score (P = 0.039). The three pairwise comparisons of the groups with regards to the systematic effect of study group that was found indicated that children born following COH-IVF had a significantly lower mean variation score than children born following MNC-IVF (Table IV). Mean variation scores of children born following COH-IVF and Sub-NC were similar; the same held true for the comparison between MNC-IVF and Sub-NC children (Table IV).

For the statistical approach, we used a population-averaged or marginal mixed model under the assumption of normality. However, the residuals of the marginal mixed models indicated that the assumption of normality was violated for the variation score, which is most likely caused by a number of children that reached the maximum score (ceiling effect). To investigate whether the conclusion on the study group effect on the variation score could still be accepted, the raw test scores were also analysed as binomial counts using generalized estimating equations (GEE) with a logit link function. The child was selected as a cluster variable with an unstructured working correlation for the three age moments. A systematic effect of study group was significant with the type III score statistic (P = 0.033). This analysis seems to confirm the marginal mixed model analysis under the assumption of normality.

2

40

Total IMP scores and other domain scores at 4, 10 and 18 months The mean total IMP and domains scores with their 95% confidence intervals are listed in Table III. The mixed-effects model indicated no overall effect of study group on the total IMP score (P = 0.066) or the other domain scores (variability score: P = 0.335, fluency score: P = 0.618 and performance score P = 0.200). The pairwise comparisons of the groups with regards to a systematic effect of study group, suggested that children born following COH-IVF had a lower mean total IMP score than children born following MNC-IVF (Table IV). However, since no overall study group effect was demonstrated, we applied Bonferroni correction to correct for the three pairwise comparisons. As a result the difference failed to

COH-IVF MNC-IVF Sub-NCn = 68 n = 57 n = 90

Male gender, n (%) 36 (53) 27 (47) 46 (51)First born, n (%) 47 (69) 38 (67) 55 (61)Birth characteristics

Gestational age in weeks, median (range) 39.4 (33-42)* 40.1 (35-43) 40.0 (30-43)*Preterm birth (< 37 weeks), n (%) 7 (10) 6 (11) 7 (8)Birthweight in grams, median (range) 3378 (1980-4700)* 3400 (2170-4680) 3565 (1150-4710)*Low birthweight (< 2500 gram), n (%) 3 (4) 4 (7) 5 (6)Small for gestational agea, n (%) 0 3 (5) 2 (2)Caesarean section, n (%) 17 (25) 8 (14) 24 (27)Signs of fetal distressb, n (%) 20 (29) 16 (28)* 40 (44)*Pregnancy-induced hypertension, n (%) 8 (12) 3 (5)* 15 (17)*

Neonatal characteristicsd

Apgar score 5 min < 7, n (%) 0 0 1 (1)Neonatal intensive-care admission, n (%) 1 (2) 2 (4) 7 (8)Breastfed for > 6 weeks, n (%) 30 (46) 26 (46) 42 (48)

Parental characteristicsd

Maternal age at conception, median (range) 32.5 (26-41) 32.8 (25-37) 33.2 (22-40)Paternal age at conception, median (range) 35.7 (28-56) 34.4 (28-48) 35.4 (25-53)Education level mother (highc), n (%) 22 (32) 22 (39) 41 (46)Education level father (highc), n (%) 29 (45) 19 (34) 33 (37)

Fertility parametersd

Intracytoplasmic sperm injection, n (%) 43 (63) 29 (51) n.a.Time to pregnancy in years, median (range) 4.1 (0-13)*** 3.8 (0-13)** 2.1 (0-11)***/**Vanishing twins, n (%) 8 (12)*/** 1 (2)* 0**

Corrected age at examination4 months (in weeks), median (range) 18 (14-23) 18 (17-21) 18 (14-21)10 months (in weeks), median (range) 44 (42-56) 44 (39-48) 44 (41-51)18 months (in years), median (range) 1.5 (1.4-1.8) 1.5 (1.4-1.7) 1.5 (1.4-1.7)

* P < 0.05.** P < 0.017 (Bonferroni correction applied).*** P < 0.001.

TABLE II. Demographic characteristics of parents and children in the Groningen ART cohort study.

Note: Mann-Whitney U -tests and Chi-square or Fisher's exact tests were used to compare groups.

c University education or vocational colleges.d Missing values for variables in the COH-IVF, MNC-IVF and Sub-NC group: < 3, <2 and <4, respectively.

a Birthweight for gestational age is < -2 standard deviation scores compared to a Dutch reference population (Dutch reference tables, Perinatal Registration Netherlands).b Signs of fetal distress denoted by meconium stained amniotic fluid and/or cardiotocographic signs and/or acidosis.

Characteristics

COH-IVF: controlled ovarian hyperstimulation IVF or ICSI, MNC-IVF: modified natural cycle IVF or ICSI, and Sub-NC: natural conception in subfertile couples.

2


41

reach statistical significance. No group effect was found in the pairwise comparisons of the variability score, the fluency score and the performance score (Table IV).

The marginal mixed effect models could not be fitted for the symmetry score, because of strong ceiling effect, as the majority of infants moved symmetrically, which caused numerical issues in the statistical analysis. At 4, 10 and 18 months of age, respectively 86.1%, 99.1%, and 98.1% of the children scored a maximum of 100% (Table III). We therefore decided not to explore this variable further and to exclude this variable from the statistical analysis. The total IMP score, the variability score, the fluency score and the performance score had no or virtually no ceiling effects.


Outcome at 4 months, mean [95% CI]Primary outcome

Variation score 91.6 [89.6 ; 93.5] 94.8 [93.2 ; 96.4] 94.5 [93.2 ; 95.7]

Additional outcomesTotal IMP scorea 77.7 [77.0 ; 78.4] 79.3 [78.5 ; 80.1] 79.1 [78.4 ; 79.8]

Variability score 71.8 [70.8 ; 72.9] 72.1 [ 71.1 ; 73.0] 71.4 [70.6 ; 72.2]

Symmetry score 99.0 [98.4 ; 99.7] 99.1 [98,4 ; 99.7] 99.4 [99.1 ; 99.8]

Fluency score 76.1 [74.8 ; 77.5] 78.5 [76.3 ; 80.8] 79.2 [77.3 ; 81.2]

Performance score 49.9 [48.8 ; 50.8] 51.5 [50.3 ; 52.7] 50.7 [49.5 ; 51.8]

Total number of children tested 67 56 86Outcome at 10 months, mean [95% CI]Primary outcome

Variation score 98.2 [97.4 ; 99.0] 98.3 [97.5 ; 99.2] 98.4 [97.7 ; 99.2]

Additional outcomesTotal IMP scoreb 90.3 [89.6 ; 91.0] 90.3 [89.5 ; 91.2] 90.9 [90.2 ; 91.7]

Variability score 82.9 [81.7 ; 84.0] 83.1 [81.7 ; 84.4] 83.7 [82.5 ; 85.0]

Symmetry score 99.9 [99.8 ; 100.0] 99.9 [99.7 ; 100.0] 100.0 [99.9 ; 100.0]

Fluency score 94.9 [93.0 ; 96.8] 94.7 [92.6 ; 96.7] 96.2 [94.5 ; 98.0]

Performance score 75.7 [74.0 ; 77.4] 76.0 [74.2 ; 77.7] 76.3 [74.8 ;77.8]

Total number of children tested 66 56 89Outcome at 18 months, mean [95% CI]Primary outcome

Variation score 96.3 [95.4 ; 97.1] 98.1 [97.5 ; 98.7] 97.0 [96.1 ; 97.8]

Additional outcomesTotal IMP scoreb 90.8 [90.1 ; 91.5] 91.7 [91.1 ; 92.3] 91.3 [90.8 ; 91.9]

Variability score 86.9 [85.6 ; 88.3] 88.9 [87.7 ; 90.2] 88.1 [87.0 ; 89.3]

Symmetry score 99.9 [99.8 ; 100.0] 99.9 [99.6 ; 100.0] 99.9 [99.7 ; 100.0]

Fluency score 83.5 [81.4 ; 85.5] 83.4 [81.5 ; 85.3] 82.8 [81.2 ; 84.4]

Performance score 87.5 [86.4 ; 88.6] 88.1 [87.2 ; 89.1] 88.7 [88.1 ; 89.4]

Total number of children tested 66 56 86

b Total IMP score originates from all five domain scores.

Note: COH-IVF: infants born following controlled ovarian hyperstimulation IVF or ICSI, MNC-IVF: infants born following modified natural cycle IVF or ICSI, and Sub-NC: naturally conceived infants born to subfertile parents.

TABLE III. Total IMP and domain scores of children in the Groningen ART cohort study.

a Total IMP score originates from four domain scores, i.e. all domains except the variability domain.

2

42

DISCUSSION

Potential effects of ovarian hyperstimulation and the in vitro procedure The present study demonstrated that children born following COH-IVF have similar scores on the IMP domain variation, the other IMP domains, and the total IMP as Sub-NC children. In addition, we found that COH-IVF children, on average, had significantly lower variation scores throughout infancy than children born following MNC-IVF. The latter two groups did, however, not differ in the other domain scores and, after Bonferroni correction, in the total IMP score.

The finding that the three groups did not differ in the total IMP score and in all but one of the domain scores is in accordance with our previous observations, which showed no differences in neurological condition in terms of minor neurological dysfunction and optimality scores between the three groups up until 2 years.135,140,172 The present study, however, found a difference in variation score between the COH-IVF group and MNC-IVF group. This underscores the notion that variation in motor behaviour taps on other neural circuitries than traditional neurological signs.174 Variation in motor behaviour in particular reflects integrity of cortical connectivity.145,150

The relatively low scores of COH-IVF children may suggest that ovarian hyperstimulation is associated with reduced movement variation throughout infancy, and that the in vitro procedure most likely is not. A potentially adverse effect of ovarian hyperstimulation could be mediated by an altered ovarian and intra-uterine environment.40 Several animal studies demonstrated that ovarian hyperstimulation is associated with alterations in hormonal levels and disturbances in the endometrium due to, among other things, a non-optimal expression of endometrial growth factors41 and reduced endometrial receptivity.41,42 The altered early environment could induce a non-optimal development of the oocyte and embryo. The latter corresponds to observations suggesting that neonatal outcomes of children born following IVF or ICSI with frozen embryo transfer are better than those of children born following IVF or ICSI with fresh embryo transfer.175,176 In case of fresh embryo

Primary outcome, mean difference [95% CI]Variation score -1.010 [-1.766 ; -0.254] 0.440 [-0.344 ; 1.224] -0.570 [-1.385 ; 0.246]

Additional outcomes, mean difference [95% CI]Total IMP score -0.787 [-1.465 ; -0.109] 0.232 [-0.442 ; 0.885] -0.556 [-1.205 ; 0.093]Variability score -0.742 [-1.735 ; 0.251] 0.468 [-0.491 ; 1.428] -0.274 [-1.227 ; 0.679]Fluency score -0.676 [-2.328 ; 0.975] -0.025 [-1.760 ; 1.709] -0.702 [-2.289 ; 0.886]Performance score -0.819 [-1.848 ; 0.211] 0.034 [-0.959 ; 1.028] -0.784 [-1.711 ; 0.202]

COH-IVF: infants born following controlled ovarian hyperstimulation IVF or ICSI, MNC-IVF: infants born following modified natural cycle IVF or ICSI, and Sub-NC: naturally conceived infants born to subfertile parents.

TABLE IV. Mixed effects model analyses on the total IMP and domain scores across infancy in the Groningen ART cohort study.

Note that the overall study group effects were adjusted for sex, gestational age, birthweight, time to pregnancy, fetal distress, vanishing twins and pregnancy-induced hypertension.

COH-IVF versus MNC-IVF MNC-IVF versus Sub-NC COH-IVF versus Sub-NC

2


43

transfer the developing embryo is exposed to a hormonally loaded endometrium. However, an argument against an unfavourable effect of ovarian hyperstimulation is the finding that the variation scores of children born following COH-IVF did not differ significantly from that of Sub-NC children. Many other studies also conclude that early neurodevelopmental outcome of children born after IVF or ICSI – i.e. the commonly applied IVF or ICSI with ovarian hyperstimulation – does not differ from children conceived naturally.89,120,177 Moreover, it should be kept in mind that the women who underwent MNC-IVF or MNC-ICSI fulfilled specific selection criteria with regards to age, BMI, previous ART treatment and the presence of an ovulatory menstrual cycle.136,137 We applied multivariable statistics to reduce selection bias, but it is conceivable that subtle group characteristics continued to affect the IMP outcomes.

Strengths and limitations The construction of our cohort has enabled us to study several components of assisted conception, in particular the effect of ovarian hyperstimulation and the in vitro procedure and the combined effect. Our control group consisted of children born to subfertile parents of whom the characteristics closely resembled those of the two IVF/ICSI groups. This resulted in minimization of confounders such as maternal age and parity, and a minimization of potential overestimations of effect.

Another strength of this study is the application of a sensitive and age-specific instrument to study neurological development. The IMP measures neurological development in terms of the quality of spontaneous motor behaviour, which is a good predictor for neurodevelopmental outcome.145 A limitation of the IMP, however, is its lack of norm scores. However, the use of non-normalized scores to study differences between groups is justified. The IMP scores of our study groups seem to be relatively high, compared to the outcomes of Heineman et al. and reflect the relatively healthy condition of the children’s nervous system.154 The latter corresponds to earlier neurological findings of our study groups.135,140,172

The prospective design of the study reduced potential selection bias – based on the child’s development or health – since we invited parents in the third trimester of their pregnancy. The longitudinal design of our study enabled us to study neurological development over time up until 18 months. Additional strengths of the present study are the blinding of the assessors to the mode of conception and the minimal postnatal attrition rate (approximately 3%).

There are several limitations to the present study. First, the study’s sample size had been calculated for neurological outcome at 18 months according to Hempel.140 We performed a post hoc sample size calculation for the variation domain score, based on a formula described by Diggle et al., using three time points, the observed effect sizes and variance-covariance estimates from the statistical analysis.178 To demonstrate effect sizes of more than 0.7 for the variation domain score between two study groups with 80% power and 5% significance level, at least 75 to 225 children had to be included per study group depending

2

44

on the covariance parameters. This means that the current study with approximately 70 subjects per study group is somewhat underpowered to detect such effects.

The lack of a fertile control group precluded conclusions on potential effects of underlying subfertility problems. The latter are known to be associated with worse obstetric, perinatal59-62 and neurodevelopmental outcome.97,98

It is well known that controlled ovarian hyperstimulation IVF/ICSI is associated with multiple births, which in turn is associated with developmental adversities.179 Therefore, developmental outcome of singletons and twins has to be studied separately. Our study was limited to singletons only, implying that the results cannot be generalized to multiples.

From a statistical point of view it should be noted that the assumption of normality was violated for the variation score, as indicated by the residuals of the marginal mixed models. This was most likely caused by ceiling effects, but not to such an extent that it hampered the analyses. A binominal model would have incorporated a ceiling effect, however, it has fewer options in choosing an appropriate covariance structure that may depend on other factors. Furthermore, the residuals of the marginal mixed models indicated that the assumption of normality was not violated with the total IMP score, the variability score, the fluency score and the performance score. For the symmetry score, however, we were unable to fit the marginal mixed-effects model because of a strong ceiling effect, and we therefore decided to exclude the variable from the statistical analysis. An effect of study group on the symmetry score would have been, however, highly unlikely, since the mean scores of the three groups are very similar (Table III).

We conclude that we observed in the present longitudinal, prospective, assessor-blinded cohort study a difference in variation score between COH-IVF and MNC-IVF children, with the latter group performing better than the former. Simultaneously, a similar difference was not observed between the two IVF groups and the Sub-NC group. Most likely these results do not imply that ovarian hyperstimulation is associated with a reduction in movement variation throughout infancy. As the application of assisted reproduction is steadily expanding and previous studies showed that reduced variation during infancy is associated with the development of neurodevelopmental disorders145,151 and reduced intelligence152, evaluation of the effect of ovarian hyperstimulation on long-term developmental outcome is warranted.

ACKNOWLEDGEMENTS

We thank participating parents and children for their cooperation and enthusiasm during the assessments; Arend F. Bos, M.D., Ph.D., for his help in initiating the study and Maaike Haadsma, M.D., Ph.D., for her help in including the participants, Inonge Reimert, M.Sc., Jolien Snijders, M.Sc., and Aïsha Sinot, M.D., for their support in collecting the data; Michiel Schrier, M.Sc., Linze Jaap Dijkstra B.Sc., and Loes de Weerd for technical assistance.

2

The Groningen ART cohort study: the effects of ovarian hyperstimulation and the IVF laboratory procedures on neurological condition at 2 years

45

CHAPTER 3 The Groningen ART cohort study: the effects of ovarian hyperstimulation and the IVF laboratory procedures on neurological condition at 2 years

Pamela Schendelaar

Karin J. Middelburg

Arend F. Bos

Maas Jan Heineman

Marjolein Jongbloed-Pereboom

Mijna Hadders-Algra

Human Reproduction 2011 ; 26 : 703 – 712

46

ABSTRACT

Background Up to 4% of children are born following assisted reproductive techniques (ART) yet relatively little is known on neurodevelopmental outcome of these children after 18 months of age. Only a limited number of long-term follow-up studies with adequate methodological quality have been reported. Our aim was to evaluate the effects of ovarian hyperstimulation, IVF laboratory procedures and a history of subfertility on neurological condition at 2 years. Methods Singletons born after controlled ovarian hyperstimulation IVF (COH-IVF, n=66), modified natural cycle IVF (MNC-IVF, n=56), natural conception in subfertile couples (Sub-NC, n=87) and in fertile couples (reference group, n=101) were assessed (using Hempel approach) by neurological examination at 2 years of age. This resulted in a neurological optimality score (NOS), a fluency score and the prevalence of minor neurological dysfunction (MND). Primary outcome was the fluency score, as fluency of movements is easily affected by subtle dysfunction of the nervous system. Results Fluency score, NOS and prevalence of MND were similar in COH-IVF, MNC-IVF and Sub-NC children. However, the fluency score (P < 0.01) and NOS (P < 0.001) of the three subfertile groups were higher, and the prevalence of MND was lower (P = 0.045), than those in the reference group. Conclusions Neurological condition of 2-year-olds born after ART is similar to that of children of subfertile couples conceived naturally. Moreover, subfertility does not seem to be associated with a worse neurological outcome. These findings are reassuring, but we have to keep in mind that subtle neurodevelopmental disorders may emerge as children grow older.

3


47

INTRODUCTION Worldwide, already 3 million children have been born following assisted reproductive techniques (ART).180 This means that even subtle adverse effects of ART on neurodevelopmental outcome will have consequences for society at large.

A recent systematic review indicated that ART does not seem to affect short-term neurodevelopmental outcome.89 This review also revealed that surprisingly little is known on long-term outcome of ART owing to the relative paucity of long-term follow-up studies of good methodological quality. Only seven good quality studies of children older than 18 months were included in the review.78,99,100,102-104,113 Since that time, only one other good quality study has been published.105 In general, the follow-up studies indicated that neurodevelopmental outcome of ART children is similar to that of children conceived naturally. However, some studies reported inconsistent results. For instance, Knoester et al. indicated that cognitive function of ICSI children was worse than that of children born following IVF and natural conception.113 Levy-Shiff et al. reported an excess of behavioural problems in IVF boys,78,100,104,105 whereas Morin et al. reported more vocalization and higher energy levels in IVF children compared with naturally conceived peers.99 Neurodevelopmental outcome after ART has also been assessed by means of register based studies. These studies suggested that IVF/ICSI per se does not increase the risk for severe cognitive impairment or neurological handicap, such as cerebral palsy.89 A disadvantage of register based studies, however, is that minor effects of ART on neurological condition cannot be evaluated.

Altogether, this means that we are currently still not well informed about potentially long-term sequelae of ART on neurological outcome. The finding that ART is associated with perinatal complications such as preterm birth and low birthweight,52-54 suggests that ART might be associated with a minor negative impact on more advanced neurodevelopmental functions.

In theory, different components of ART may affect development. Hypothetical points of concern are ovarian hyperstimulation,171 the impact of the ART procedure itself31 and consequences of vanishing twins.50 In addition, background factors associated with ART, such as a history of subfertility resulting from tuba pathology, andrological factors, or increased maternal age,57 may contribute to worse neurodevelopmental outcome in ART singletons.55,59

To examine potential effects of the various components of ART, the Groningen ART cohort was composed (Figure I).135,140 The cohort consists of three groups. The first group consisted of children born after conventional controlled ovarian hyperstimulation IVF (COH-IVF), the second of children born after IVF in the modified natural cycle (MNC-IVF) in which medication use was minimal,181 and the third group consisted of a control group of naturally conceived children born to subfertile couples (Sub-NC). Possible differences in outcome of COH-IVF and MNC-IVF children may be attributed to ovarian hyperstimulation, whereas

3

48

potential differences in MNC-IVF and Sub-NC children may largely be attributed to the ART procedure.

We previously reported that neurological outcome of children of the Groningen ART cohort was similar at 2 weeks and 3 months of age and up to and including 18 months.135,140 However, neurological condition at 3 months of the Groningen ART cohort was less favourable than that of a reference group of the general population, suggesting that neither ovarian hyperstimulation, nor the in vitro procedure, but rather factors associated with subfertility affect early neurodevelopmental outcome.135

The primary aim of this paper is to assess the effect of ovarian hyperstimulation and the IVF laboratory procedure on neurological outcome at 2 years. To this end, all children of the prospective Groningen ART cohort were reassessed using the standardized, precise and age-specific neurological assessment according to Hempel.182 Age-specific testing is necessary, since many functional and structural changes of the brain occur during childhood. Primary outcome was the neurological condition expressed in terms of fluency of motor behaviour. This is a sensitive measure to detect subtle changes in neuromotor development, since minor dysfunction of the nervous system already results in a reduction of fluency of motion.161,162 Secondary outcome parameters were the neurological optimality score (NOS) and type and severity of minor neurological dysfunction (MND). Specific attention was paid to relationships between causes of subfertility and neurological outcome. The secondary aim of the study was to compare neurological outcome of the Groningen ART study with that of a newly recruited retrospective reference group of 2-year-olds born to fertile couples.

FIGURE I. The Groningen ART cohort study. The effects of ovarian hyperstimulation, the IVF laboratory procedure, the combination of both and a history of subfertility are studied by four different comparisons. The groups controlled ovarian hyperstimulation IVF (COH-IVF), modified natural cycle IVF (MNC-IVF) and natural conception in subfertile couples (Sub-NC) were pooled to form the subfertile group.

3


49

MATERIAL AND METHODS

Participants Pregnant subfertile couples with a term date between March 2005 and December 2006 were recruited at the Department of Reproductive Medicine of the University Medical Center Groningen.135 All couples who achieved a singleton pregnancy following IVF or ICSI were invited to participate, resulting in groups of children born after COH-IVF and MNC-IVF. Couples with a pregnancy after treatment with cryopreserved or donated oocytes or embryos were excluded. The third group formed was Sub-NC, which comprised couples that had tried to conceive for at least 1 year, and finally conceived naturally while waiting for fertility evaluation or treatment. We restricted our analysis to singletons, as – in contrast to COH-IVF – MNC-IVF rarely results in multiple gestation, and being a member of a multiple is associated with an increased risk for developmental problems.179

For the present study, a new retrospective reference group was recruited between February and October 2009 at six child welfare clinics in and around Groningen. All parents of 2-year-old children that visited the child welfare clinic for routine general health care were invited to participate. Children of couples who had attempted to achieve pregnancy for more than 1 year or achieved pregnancy by any form of assisted conception were excluded.

Prenatal, perinatal and demographic information had been gathered by use of standardized charts during the first follow-up assessment, ~2 weeks post-term.135 When information was incomplete or complications during pregnancy or birth had occurred, midwifes and gynaecologists were asked for details. Detailed information on the causes and treatment of infertility was retrieved from medical records. The Medical Ethical Commission of the University Medical Center Groningen approved the study design. At least one of the children’s parents provided written informed consent for participation of their child in the study.

Neurological assessment All children were assessed around the time of their second birthday. Neurological assessment was carried out according to Hempel: this assessment is a standardized neurological examination to assess MND at preschool age.182 Five domains of function were assessed which can be scored as typical or deviant: fine motor function, gross motor function, posture and muscle tone, reflexes and visuomotor function (denoting function of the visual system and eye movements).162

Children were classified as neurologically normal, simple MND, complex MND or neurologically abnormal. Simple MND denotes the presence of one deviant domain, and is regarded as a non-optimal, yet normal form of brain function.90 Complex MND means the presence of more than one domain of dysfunction and represents the clinically relevant form, as it is associated with preterm birth and perinatal adversities, and behavioural and learning disorders.157,162 Neurologically abnormal implies the presence of a distinct

3

50

neurological syndrome, such as cerebral palsy. Neurologically normal implies the absence of neurological dysfunction and is scored when no domains are deviant or only the domain of reflexes is deviant.

In addition, outcome of the Hempel assessment was expressed in a NOS. The NOS consists of 58 items, for which an optimal condition is defined (range 0- 58). The sum of the items which fulfil the criteria for optimality forms the score. Higher scores represent better performance. It is important to realize that there is a conceptual difference between normality and optimality, since the range for optimal behaviour is narrower than for normal behaviour.160 This implies that both scores are suitable tools to evaluate subtle differences in neurological outcome. The fluency score (range 0- 13), a sub-score of the NOS, deals with the fluency of motor behaviour. Since subtle dysfunction of the nervous system is most easily expressed in a reduction of the fluency of movements, this measure is the most sensitive one to detect minimal changes in neuromotor development.161,162 The Hempel assessment has a satisfactory inter-rater reliability [κ = 0.62-1.00 (mean 0.93)] and a good construct validity.155 Note that in the Results section neurological outcome is first presented in the clinical terms of MND. Thereafter, the results of the primary outcome parameter, the fluency score, and the other additional outcome parameter, the NOS, are described.

M.J.P. and K.J.M., who assessed the children of the subfertile groups, were unaware of the mode of conception. Parents were asked not to reveal information about the conception method. However, the reference group was recruited separately from the subfertile groups, therefore it was impossible to keep the assessor of this group (M.J.P.) blind.

Statistical analysis Power calculation was based on neurological outcome at 18 months. To detect at least half a SD difference in the fluency score (mean 9.5, SD 1.7),161 with 80% power at least 64 children had to be included per group.140

Chi-square test, Fisher’s exact test, Mann-Whitney U-test or Student’s t-test were used to test differences between groups. Potential associations between group status and neurological outcome were analysed using regression analyses. Linear regression analysis was used for the fluency score and NOS: to this end, both scores were normalized using –log (14.5-fluency score) and -ln (59.5-NOS), respectively. Logistic regression analysis was applied in the analysis of group effects on the occurrence of MND. A similar set of analyses was applied to assess the effect of specific causes of subfertility on neurological outcome.

In the multivariable analyses, background variables were included for which the groups differed at 5% significance level, with the exception of the subfertility causes. In addition, gestational age was entered in the multivariable analyses on an a priori basis, as gestational age is tightly linked to neurological outcome.

The results of the multiple linear regression analyses were used to calculate confidence intervals (CIs) for adjusted differences between the means of the groups. To interpret these on the original scale, we use the fact that the difference between means of two groups, A

3


51

and B, on the transformed scale for the fluency score can be interpreted as the common logarithm of the ratio (14.5 – medB) / (14.5 – medA), where medA and medB are medians on the original scale.

Statistical analyses were performed using the Statistical Package for the Social Sciences 15.0 for Windows. Bonferroni corrections were applied in the analyses of the outcome parameters, but not in the analyses of patient characteristics. P-values of 5% or less were considered significant.

RESULTS

Participation and demographic characteristics Eighty-nine children born after COH-IVF, 79 MNC-IVF children and 143 Sub-NC children met the inclusion criteria during the prenatal period: parents of 68 (76%), 57 (72%) and 90 (63%) children, respectively, agreed to participate in neurodevelopmental follow-up (Figure II).135 Overall, obstetrical, neonatal and social characteristics of participants and non-participants were similar, except for maternal age.135

Six children were lost to follow-up at the assessment at 2 years. Two COH-IVF children and three Sub-NC children did not participate for logistical reasons. One MNC-IVF girl had a congenital heart disorder and died at 3 weeks of age.

Overall, demographic characteristics of the three groups were similar. The differences found were the following: gestational age was shorter in COH-IVF children than in Sub-NC children (P = 0.012). COH-IVF children were more often survivors of a vanishing twin than MNC-IVF and Sub-NC children (P = 0.038 and P < 0.001, respectively). Mothers of COH-IVF and MNC-IVF children needed more time to get pregnant than mothers of Sub-NC children (P < 0.001 and P = 0.003, respectively). Finally, fathers of MNC-IVF children were younger than those of COH-IVF children (P = 0.037) (Table I).

Neurological condition of children of the Groningen ART cohort None of the children of the COH-IVF, MNC-IVF and Sub-NC groups showed a definitely abnormal neurological condition, and none had an ICD-10 neurodevelopmental diagnosis. Children of the COH-IVF group tended to show both simple and complex MND more often than MNC-IVF or Sub-NC children, but this neurological disadvantage did not reach statistical significance (Table IIa). In accordance, the prevalence of dysfunction in specific neurological domains was similar in the three groups. As the prevalence of complex MND was low (Table II), we used in the multivariable statistical analysis MND, denoting the presence of either simple or complex MND, as outcome parameter. The multivariable analyses confirmed that neither the ovarian hyperstimulation (COH-IVF versus MNC-IVF), nor the in vitro laboratory procedures (MNC-IVF versus Sub-NC), nor the combination of both factors (COH-IVF versus Sub-NC affected the occurrence of MND (Table III).

3

52

The primary outcome parameter, the fluency score, and the other additional outcome parameter, the NOS, were similar in the three groups (Figure III). The similarity in fluency score and NOS in the three groups was confirmed in the multivariable analyses (Table III)

Transforming the CI for the differences between the groups (Table III) back into the original scale results in the following interpretation: assuming that the corrected median fluency score is 10 in the Sub-NC group, the CIs for corrected medians for group MNC-IVF and COH-IVF are both (9.7 to 10.4); assuming that the score is 10 for the MNC-IVF group, the CI for the COH-IVF group median is (9.6 to 10.4). For the NOS, assuming that the median score of 49 for the Sub-NC group results in CIs (47.5 to 50.1) and (46.8 to 49.7) for medians in the group MNC-IVF and COH-IVF, respectively; assuming the score is 48 for the COH-IVF group, the CI for the MNC-IVF group is (45.7 to 46.9). The resemblance in neurological condition of the three groups allowed us to pool them to form the subfertile group.

Figure II. Flow chart of the Groningen ART cohort.

3


53

Neurological condition of the subfertile and reference groups Parents of 167 reference children were invited to participate in the study. Parents of 37 infants refused to participate. Logistic problems hampered the assessment of another 29 children. Eventually, 101 (61%) children were examined. Maternal education of non-participating children was significantly lower than that of participating children (P < 0.01), but gender distribution and gestational age at birth were similar for participants and non-participants. Table IV illustrates that perinatal and social characteristics of the subfertile and


Male gender, n (%) 36 (55) 27 (48) 45 (52)First born, n (%) 45 (68) 38 (68) 53 (61)Birth characteristics

Gestational age in weeks, median (range) 39.4 (33.4-42.3)* 40.1 (34.6-42.6) 40.0 (30.1-42.7)*Preterm birth (< 37 weeks), n (%) 7 (11) 6 (11) 7 (8)Birthweight in grams, mean (SD) 3396 (550) 3400 (576) 3547 (587)Low birthweight (< 2500 gram), n (%) 3 (5) 4 (7) 5 (6)Small for gestational agea, n (%) 0 3 (5) 2 (2)Caesarean section, n (%) 16 (24) 8 (14) 23 (26)Signs of fetal distressb, n (%) 19 (29) 16 (29) 38 (44)

Neonatal characteristicsApgar score 5 min < 7c, n (%) 0 0 1 (1)Neonatal intensive care admission, n (%) 1 (2) 2 (4) 6 (7)Breasfed for > 6 weeksc, n (%) 30 (48) 26 (46) 42 (49)

Parental characteristicsMaternal age at conception in years, median (range) 32.9 (26.3-40.9) 32.8 (25.3-37.5) 33.3 (22.2-40.3)Paternal age at conception in yearsc, median (range) 35.7 (27.5-56.1)* 34.4 (28.3-47.8)* 35.4 (25.5-52.6)Smoking during pregnancy, n (%) 7 (11) 7 (13) 9 (10)

Parental socioeconomic statusEducation level mother (highd), n (%) 22 (33) 21 (38) 40 (46)Education level father (highd)c, n (%) 29 (46) 18 (32) 32 (47)

ICSI performed, n (%) 42 (64) 28 (50) n.a.Vanishing twins, n (%) 8 (12)*/*** 1 (2)* 0 (0)***Time to pregnancy in yearsc, median (range) 4.0 (0.1-13.3)*** 3.8 (0.1-13.2)** 2.1 (0.1-11.3)***/**Type of infertility (primary), n (%) 35 (53) 33 (59) 45 (52)Subfertility causese

Tuba pathology, n (%) 8 (15) 9 (17) 5 (6)Male factor, n (%) 28 (52)** 31 (57)*** 19 (24)**/***Other causes, n (%) 9 (17) 5 (9) 9 (11)Unknown cause, n (%) 8 (15)*** 9 (17)*** 48 (59)***/***

Corrected age at examination at 2 years of age (in months); median (range) 24.9 (23.3-30.2) 24.9 (13.8-27.9) 25.0 (23.2-28.9)

b Signs of fetal distress denoted by meconium stained amniotic fluid and/or cardiotocographic signs and/or acidosis.

d University education or vocational colleges.

COH-IVF: controlled ovarian hyperstimulation IVF, MNC-IVF: modified natural cycle IVF, Sub-NC: natural conception in subfertile couples.

TABLE I. Characteristics of parents and infants of the Groningen ART cohort.

Note: Mann-Whitney U -tests or Student's t -test and Chi-square tests or Fisher's exact tests were used to compare between groups; * P < 0.05; ** P < 0.01, *** P < 0.001.

Characteristics

e Couples may have more than 1 cause of subfertility, therefore totals may exceed 100% .

c Missing data in three groups: Apgar n=5, breastfed n=5, paternal age n=4, education level father n=4, time to pregnancy n=1. Note that in the percentages missing values have been taken into account.

a Birthweight for gestational age is < -2 standard deviations compared with the Dutch reference population (Dutch reference tables, perinatal Registration Netherlands).

3

54

the reference groups differed considerably. Compared to the fertile reference group children of the subfertile group more often were firstborn (P = 0.022) and preterm (P = 0.015), had a lower birthweight (P = 0.022), showed more often signs of fetal distress (P < 0.001), more often were born following Caesarean section (P = 0.002) and were slightly younger at the follow-up assessment (P < 0.001). In addition, in the subfertile group maternal age was higher (P < 0.001) and parental educational level was lower (maternal: P = 0.034, paternal: P = 0.004) than in the reference group. Finally, parents of the subfertile group needed more time to get pregnant (P < 0.001).

Children of the subfertile group tended to show less MND than the reference children. In the univariable analysis the difference did not reach statistical significance (Table IIa). Yet, when confounders were taken into account the difference was statistically significant (P = 0.045; Table III). Furthermore, a minor difference was found in the specific domains of dysfunction: the reference group showed more often dysfunctional posture and muscle tone regulation than the subfertile group (P < 0.01; Table IIa).

Similar results were found for the fluency score and the NOS. Both fluency score and NOS of the subfertile group were higher than those of the reference group [fluency score: median values 10 (6 to 12) and 9 (6 to 13), respectively; P = 0.002; NOS: median values 49 (32 to 57) and 47 (28 to 55); P < 0.001; Figure III]. Multivariable analyses confirmed that differences in fluency and NOS scores between the subfertile group and the reference group were statistically significant (P = 0.003 and P = 0.001, respectively; Table III).

Transforming the CIs for the differences between both groups (Table III) back into the original scale results in the following interpretation: assuming that the corrected median fluency score is 9 in the reference group, the CIs for corrected medians for the subfertile group is (9.2 to 9.8); assuming that the score is 10 for the subfertile group, the CI for the reference group median is (10.1 to 10.6). For the NOS, assuming that the median score of 47 for the reference group results in the CI (47.8 to 49.7) for the median in the subfertile group; assuming that the score is 49 for the subfertile group, the CI for the reference group is (49.6 to 51.3).

Underlying subfertility causes and neurological outcome None of the specific causes of subfertility were related to the fluency score, NOS or the type or severity of MND (Table IIb). This was confirmed in multivariable analyses (data not provided).

3


55

3

COH – IVF MNC – IVF Sub – NC

groupa Reference groupb

n = 66 n = 56 n = 87 n = 209 n = 101

Normal 58 (88%) 52 (93%) 83 (95%) 193 (92%) 88 (87%)Simple MNDc 5 (8%) 3 (5%) 2 (2%) 10 (5%) 10 (10%)

Complex MNDc 3 (5%) 1 (2%) 2 (2%) 6 (3%) 3 (3%)

Fine motor dysfunction 0 1 (2%) 0 1 (1%) 0Gross motor dysfunction 4 (6%) 1 (2%) 3 (3%) 8 (4%) 3 (3%)

Posture and muscle tone dysfunction 4 (6%) 2 (4%) 1 (1%) 7 (3%)* 12 (11%)*Dysfunctional reflexes 11 (17%) 9 (16%) 12 (14%) 32 (15%) 22 (22%)

Visuomotor dysfunction 1 (2%) 0 0 1 (1%) 0

Tuba pathology Male factor Other causese Unknown causen = 29 n = 94 n = 40 n = 65

Neurological outcomeNormal 27 (93) 85 (90) 37 (93) 60 (92)

Simple MNDc 2 (7) 6 (6) 2 (5) 2 (3)Complex MNDc 0 3 (3) 1 (3) 3 (5)

b. Subfertility causes d

Note: Chi-square test or Fisher's Exact test were used to compare groups. * P < 0.01.

Domain of dysfunction

Neurological outcome

e Other known causes of subertility: endometriosis, cervical factor, hormonal cause, lesbian couple.

d Couples may have more than 1 cause of subfertility, therefore totals may exceed 100% .

TABLE II. Neurological classification and domains of dysfunction.a. ART cohort and reference group

c MND = minor neurological dysfunction.

b Naturally conceived children born to fertile parents.

a Group COH-IVF, MNC-IVF and Sub-NC were pooled to form the subfertile group.

Adjusted median difference (95% CI) P - value

COH-IVF versus Sub-NC 0.006 (-0.027 – 0.039)a 0.700MNC-IVF versus Sub-NC 0.007 (-0.026 – 0.040)b 0.676COH-IVF versus MNC-IVF 0.003 (-0.033 – 0.038)c 0.888Subfertile group versus reference group 0.039 (0.014 – 0.064)d 0.003

COH-IVF versus Sub-NC -0.061 (-0.187 – 0.066)a 0.344MNC-IVF versus Sub-NC -0.011 (-0.137 – 0.115)b 0.867COH-IVF versus MNC-IVF -0.045 (-0.182 – 0.093)c 0.523Subfertile group versus reference group 0.153 (0.062 – 0.245)d 0.001

Adjusted odds ratio (95% CI) P - value

COH-IVF versus Sub-NC 2.161 (0.575 – 8.122)a 0.254MNC-IVF versus Sub-NC 1.221 (0.277 – 5.393)b 0.792COH-IVF versus MNC-IVF 1.917 (0.518 – 7.100)c 0.330Subfertile group versus reference group 0.386 (0.152 – 0.979)d 0.045

TABLE III. Multivariate regression analyses of the influence of ART components and history of subfertility on neurological outcome.

Note: Linear regression analyses were used to compare the fluency score and NOS between groups. Logistic regression analyses were used to compare the prevalence of MND between groups.

d Adjusted for gestational age, birthweight, firstborn, signs of fetal distress, Caesarean section, breastfed for > 6 weeks, maternal age, education level mother, education level father and corrected age at examination at 2 years.

a Adjusted for gestational age, time to pregnancy, gestational age and vanishing twins. Bonferroni correction applied.b Adjusted for gestational age and time to pregnancy. Bonferroni correction applied.

Outcome measure: Neurological optimality score - ln (59.5 - NOS)

Outcome measure: Fluency score - log (14.5 - fluency score)

Outcome measure: MND e

e MND = simple and complex minor neurological dysfunction.

c Adjusted for gestational age, paternal age and vanishing twins. Bonferroni correction applied.

56

DISCUSSION

The present study indicates that neurological outcome in 2-year-olds is not influenced by ovarian hyperstimulation, the in vitro laboratory procedure itself or a combination of both. The findings support the results of the Groningen ART cohort study at younger ages.135,135,140

Our results are in line with most good quality studies in which children born after ART were followed until at least the age of 2 years.78,102-105 They strengthen the notion that ART is not associated with adverse neurological outcome at early age since we used a standardized and sensitive neurological assessment which allows for the detection of subtle differences in outcome.155 Our data also indicated that specific causes of subfertility, which may be associated with higher rates of spontaneous abortion,55 are not associated with worse neurological outcome.

3

FIGURE III. Fluency scores and NOS. Fluency scores and NOSs in the three ART cohort groups (upper panel) and the subfertile group and reference group (lower panel). * P < 0.01, ** P < 0.001.


57

Unexpectedly, we found that neurological outcome of the subfertile group was better than that of the reference group. Previously we found that neurological condition of the subfertile group at 3 months was slightly worse than that of a reference group of the general population.135 This difference may be attributed to the selection of the reference groups. The reference group at 3 months consisted of a group which was representative for the general population – the children had been assessed as part of a general health check-up provided for all children. The current reference group consisted of 2-year-olds whose parents volunteered to form a reference group for the ART cohort study. This may have introduced a selection bias. Also, Knoester et al. reported in the control group of naturally conceived children of fertile couples a prevalence rate of MND which was substantially higher than that in the general population.104,183 Recently, Carson et al. elegantly demonstrated the importance of a proper comparison group: they showed that when the control group was not properly matched to the ART study group, children born after ART had a better cognitive performance than the reference group.184 But when the background factors of the reference group closely resembled that of the ART group differences in cognitive disabilities disappeared.184 Perinatal and social background factors of our reference group differed substantially from those of the Groningen ART cohort groups.

3

Subfertile groupa Reference groupn = 209 n = 101

Male gender, n (%) 108 (52) 47 (47) 0.396First born, n (%) 136 (65) 52 (52) 0.022Birth characteristics

Gestational age (in weeks), median (range) 39.9 (30.1-42.7) 40.0 (35.7-42.4) 0.275Birthweight in gramsb, mean (range) 3460 (575) 3615 (465) 0.022Lowbirth weight (< 2500 gram), n (%) 12 (6) 0 (0) 0.021Small for gestational agec, n (%) 5 (2) 6 (6) 0.168Preterm birth (< 37 weeks), n (%) 20 (10) 2 (2) 0.015Signs of fetal distressd, n (%) 73 (35) 20 (20) <0.001Caesarean section, n (%) 47 (23) 8 (8) 0.002

Neonatal characteristicsBreastfed for > 6 weeksb, n (%) 98 (48) 59 (62) 0.030

Parental characteristicsMaternal age at conception, mean (range) 33.0 (22.2-40.9) 30.9 (18.0-40.4) <0.001Smoking during pregnancy (mother), n (%) 23 (11) 11 (11) 0.976

Parental socioeconomic statusEducation level mother (highe), n (%) 83 (40) 53 (53) 0.034Education level father (highe)b, n (%) 79 (39) 55 (56) 0.004

Time to pregnancy (categoricalf)b, median (range) 4 (0-5) 0 (0-1) <0.001Corrected age at examination at 2 years of age (in months), median (range) 24.9 (23.2-30.2) 25.5 (23.5-29.1) <0.001

e University education or vocational colleges.

c Birthweight for gestational age is < -2 standard deviations compared with the Dutch reference population (Dutch reference tables, perinatal Registration Netherlands).

f Time to pregnancy as a categorical value: 0 = 0 - ½ year, 1 = ½ - 1 year, 2 = 1 - 2 years, 3 = 2 -3 years, 4 = 3 - 5 years, 5 = > 5 years.

TABLE IV. Characteristics of parents and infants of the subfertile group and the reference group.

Characteristics P - value

Note: Mann-Whitney U -tests or Student's t -test and Chi-square tests or Fisher's exact tests were used to compare between groups.a Group COH-IVF, MNC-IVF and Sub-NC were pooled to form the subfertile group.

d Signs of fetal distress denoted by meconium stained amniotic fluid and/or cardiotocographic signs and/or acidosis.

b Missing data in two groups: birthweight n=7, breastfed n=10, education level father n=4, time to pregnancy n=1. Note that in the percentages missing values have been taken into account.

58

From a neurodevelopmental view perinatal and social background factors in the reference group were more favourable than those in the subfertile group (Table IV). Multivariable analyses were applied to adjust for the large differences in background between the two groups.

Strengths and limitations One of the major strengths of the study is the composition of our study groups which allowed for a disentangling of the effect of two aspects of ART. Our study was able to demonstrate that neither ovarian hyperstimulation nor the in vitro procedure was associated with an increased risk for neurological non-optimality at the age of 2 years.

Parental characteristics of the naturally conceived control group, consisting of children born to subfertile couples (Sub-NC), closely resembled those of both COH-IVF and MNC-IVF groups. As a result, the effects of potential confounders and a potential overestimation of the effect of ART were minimized.

Another major strength of the present study is the application of a sensitive and age-specific technique to assess neurological condition. The strength of the Hempel assessment is illustrated by the study of Bouwstra et al. which demonstrated a negative effect of neonatal trans-fatty acid status on neurodevelopmental outcome using the assessment according to Hempel but not by using the Bayley’s Scale of Infant Development.164 Although subtle differences in neurological outcome, such as a few points reduction in the NOS or fluency score, may not have clinical relevance for individual persons, minor deviations in neurodevelopmental outcome for substantial subpopulations may affect society at large, especially since the number of fertility problems is steadily rising.

The prospective design of our study, in which couples of the COH-IVF, MNC-IVF and Sub-NC groups were invited in the third trimester of pregnancy, reduced potential selection bias based on the child’s health or development. Additional strengths of our study are the minimal postnatal attrition (3%) and the ‘blinding’ of assessors to the mode of conception of the ART cohort. Whether blinding in ART studies is valuable, was questioned by Ludwig et al. who found that the assessors’ feeling about the mode of conception was correct in 75% of cases.185 The likelihood that the assessors in our study would guess the conception mode was minimized, because all parents of the ART cohort had experienced subfertility. However, we had been unable to prevent knowledge of conception mode of the assessor involved in the examination of the reference group.

It is a limitation of the study that we studied singletons only. This means that the results cannot be generalized to children born after multiple gestation. It is well known that controlled ovarian hyperstimulation IVF is associated with multiple birth and that being a member of a multiple is associated with an increased risk for developmental problems.179

A limitation of the study is the composition of the fertile reference group. Our results underscore the need of prospectively recruited control groups. Another limitation of the study was the size of the MNC-IVF group, which was slightly smaller than the 64 needed for

3


59

an adequate power of the study. However, the finding that outcome of the three subfertile groups was very similar, suggests that it is unlikely that larger groups would have revealed statistically significant differences between the groups. Owing to the relatively small number of children in our groups, the study was unable to evaluate the association between ART procedures and rare, severe neurodevelopmental disorders.

In conclusion, the neurological condition of 2-year-olds born after COH-IVF or MNC-IVF is similar to that of peers born to subfertile couples who conceived without ART. This indicates that neither the ovarian hyperstimulation, nor the IVF laboratory procedures or a combination of both factors is associated with a worse neurological outcome up until the age of 2 years. Additionally, our data suggest that subfertility is not associated with a worse neurological outcome at 2 years. Although the findings of the present Groningen ART cohort study are reassuring, we have to keep in mind that subtle neurodevelopmental disorders may emerge when children grow older.

ACKNOWLEDGEMENTS

We thank participating parents and children for their cooperation and enthusiasm; Maaike Haadsma, M.D., Ph.D. for her help in including the participants; Michiel Schrier, M.Sc., Linze Dijkstra B.Sc., and Loes de Weerd for technical assistance.

3

60

Increased time to pregnancy is associated with suboptimal neurological condition of 2-year-olds

61

CHAPTER 4 Increased time to pregnancy is associated with suboptimal neurological condition of 2-year-olds

Jorien Seggers

Pamela Schendelaar

Arend F. Bos

Maas Jan Heineman

Karin J. Middelburg

Maaike L. Haadsma

Mijna Hadders-Algra

Archives of Disease in Childhood, Fetal and Neonatal Edition 2013 ; 98 : F434 – 436

62

ABSTRACT Objective To evaluate whether time to pregnancy (TTP) is associated with neurological condition of 2-year-old children born to subfertile parents. Design Data collected in a prospective, assessor-blinded follow-up study were used for cross-sectional analyses. Patients Participants were the singletons of the Groningen assisted reproductive techniques (ART) cohort study: all children were born to subfertile couples (n=209). The active waiting TTP of the couples was obtained from fertility charts was recorded in years and months, and was converted to decimal years. Main outcome measure The presence of minor neurological dysfunction (MND), assessed with the Hempel examination. Results MND was present in 16 (7.7%) children. TTP of children with MND (median 4.1, range 1.6-13.2) was significantly longer than that of children without MND (median 2.8, range 0.1-13.3; Mann-Whitney U-test P = 0.014). Logistic regression analysis on the contribution of TTP to MND resulted in a crude OR of 1.27 (95% CI: 1.06-1.54). After correction for gestational age, parental age and parental level of education, the association remained statistically significant: OR=1.30 (95% CI: 1.05-1.61). Conclusion Increased TTP was associated with suboptimal neurological development in 2-year-old children. This suggests that subfertility and its determinants are involved in the genesis of neurodevelopmental problems.

4


63

INTRODUCTION It is well established that children conceived with assisted reproductive techniques (ART), like in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI), are at increased risk for preterm birth and low birthweight,53 conditions associated with developmental problems. Previously, we reported that neurological condition at 2 years was not affected by IVF/ICSI, that is, it was not affected by ovarian hyperstimulation nor by the in vitro procedure.172 An increasing body of evidence suggests that subfertility per se, rather than fertility treatment, is associated with adverse outcomes.59,65

In order to evaluate the effect of subfertility, time to pregnancy (TTP) may be used. The notion that TTP can be used as a proxy for the severity of subfertility is based on models that have shown that TTP is an important predictor of the chance of pregnancy in subfertile couples.138,139 About ten to twenty percent of couples are subfertile: they fail to achieve a successful pregnancy after 12 months or more of appropriate, timed unprotected intercourse.186 Other studies demonstrated that increased time to pregnancy is associated with more Caesarean sections, pregnancy-induced hypertension, preterm birth, perinatal death, and birth defects.60-63,66,187 Zhu et al. demonstrated that increasing TTP is correlated with a modest delay in psychomotor development in 18-month-olds children.97 However, the association between prolonged TTP and detailed neurological outcome is not known. Therefore, the aim of this study is to evaluate the association between TTP and neurological status of 2-year-old children.

METHODS This study is part of the Groningen ART cohort study, a prospective assessor-blinded longitudinal study that focuses on developmental outcome of children born following IVF and ICSI. Two hundred and nine singletons were included in the analyses: 66 children born after controlled ovarian hyperstimulation-IVF/ICSI, 56 born after modified natural cycle-IVF/ICSI and 87 children born naturally to subfertile couples. As group status did not affect neurological condition at 2 years,172 we pooled the groups to form a set of 209 singletons born to subfertile couples.

Information on TTP was obtained from fertility charts and is recorded in years and months. Subsequently, TTP was converted into decimal years. TTP was defined as the time between the active child wish (i.e. the start of timed unprotected intercourse) of the couple and conception. Note that in case of miscarriage, TTP had a new onset and ended at conception of the child included in the present study. This explains why subfertile couples may have a TTP of < 1 year.

Around the time of the second birthday of children, neurological condition was assessed using the neurological examination according to Hempel.182 In the Hempel assessment five domains of function are assessed as typical or deviant: fine motor function, gross motor

4

64

function, posture and muscle tone, reflexes, and visuomotor function. Children were classified as neurologically normal, simple minor neurological dysfunction (MND) or complex MND.90 Children are classified as neurologically normal if no domains are deviant or only the domain of reflexes. Simple MND implies the presence of one deviant domain (except reflexes) and it indicates a suboptimal, yet normal form of brain function. Complex MND implies the presence of multiple deviant domains and it represents the clinically relevant form of MND.

Details on the inclusion period and materials and methods have been reported previously.140,172 Parents gave written informed consent for participation of their child in the study. The study design was approved by the ethics committee of the University Medical Center Groningen.

Statistical analysis First, we reported background characteristics of the Groningen ART cohort in a descriptive table. Second, the Mann-Whitney U-test was performed to evaluate differences in TTP between children with or without MND. Third, multiple logistic regression analyses were used to study the effect of TTP on the occurrence of MND (either simple or complex MND) while correcting for gestational age in weeks, parental age in years and parental level of education (high or not, used as an indicator of socio-economic status). Fourth, we repeated the multiple logistic regression analysis to evaluate the effect of TTP on complex MND alone, i.e. on the clinically relevant form of MND. We used Predictive Analytics SoftWare (PASW) version 18.0 to analyse the data. Probability values of <0.05 were considered statistically significant.

RESULTS

Table I shows the background characteristics of the children assessed (n=209). MND was present in 16 (7.7%) children: simple MND occurred in 10 children and complex MND in 6 children. TTP pregnancy showed a skewed distribution, median (range): 3.0 (0.1-13.3) years. TTP of children with MND (median 4.1, range 1.6-13.2) was significantly longer than that of children without MND (median 2.8, range 0.1-13.3; Mann-Whitney U-test P = 0.014, see Figure I). Logistic regression of the contribution of TTP in years to the presence of MND (either simple or complex MND) resulted in a crude odds ratio (OR) of 1.27 (95% confidence interval (CI): 1.06-1.54; P = 0.012). After correction for gestational age, parental age and parental level of education, the association remained statistically significant: adjusted OR=1.30 (95%CI: 1.05-1.61; P = 0.018). Logistic regression of the contribution of TTP in years to complex MND resulted in similar associations: crude OR=1.38 (95%CI: 1.07-1.77; P = 0.014), adjusted OR=2.29 (95%CI: 1.03-5.09; P = 0.041).

The association between TTP and MND was mainly brought about by deviancies in the domain posture and muscle tone (n=7): crude OR=1.36 (95%CI: 1.07-1.74 P = 0.012). After

4


65

correction for gestational age, parental age and parental level of education, this association lost statistical significance: adjusted OR=1.31 (95%CI: 0.97-1.79; P = 0.081). The other domains did not show a statistically significant association with TTP.

DISCUSSION TTP was positively associated with MND, including the clinically relevant form complex MND. This finding is in line with the study of Zhu et al., who demonstrated that increasing TTP is correlated with a modest delay in psychomotor development in 18-month-old children.97 A major strength of the present study is the application of sensitive and age-specific methods to assess neurological condition. In addition, attrition was minimal (3%), and selection bias was reduced as couples were recruited prospectively during the third trimester of pregnancy. Note that we were able to pool the three original study groups to form one subfertile group as group status did not affect neurological outcome. In this way, we could increase the power of the study to detect an association between TTP and MND. We checked whether additional correction for group status in the multiple analyses altered the results; it did not (data not shown). A limitation of the study is that twins were not

Characteristcs n = 209

Male gender, n (%) 108 (52)First born, n (%) 136 (65.1)Gestational characteristics Smoking during pregnancy, n (%) 23 (11) Use of folic acid during pregnancya, n (%) 191 (91.4) Pregnancy-induced hypertension, n (%) 25 (12) Gestational diabetes, n (%) 3 (1.4)Birth characteristics Gestational age (weeks), median (range) 39.6 (1.9) Preterm birth (<37 weeks), n (%) 20 (9.6) Birthweight (g), mean (SD) 3460 (575)Parental characteristics Time to pregnancy in yearsa, median (range) 3.0 (0.1-13.3) Maternal age at conception, median (range) 33.0 (22.2-40.9) Paternal age at conceptiona, median (range) 35.1 (25.5-56.1) Education level mother highb, n (%) 83 (39.7) Education level father higha/b, n (%) 79 (37.8)

b University education or vocational colleges.

ART: assisted reproductive techniques.

a Missing values: education level father n=4, use of folic acid n=3, paternal age at conception n=4, time to pregnancy n=1.

TABLE I. Infant, gestational, birth and parental characteristics of the singletons participating in the Groningen ART cohort study.

Values are number (percentage), or mean (SD) or median (range).

Note that in case of miscarriage, time to pregnancy had a new onset and ended at conception of the child included in the present study. This explains why subfertile couples may have a time to pregnancy of < 1 year.

4

66

included. This means that the results cannot be generalized to children born after multiple gestation. Furthermore, it should be noted that only six children had complex MND, which means that caution is deserved in the interpretation of the association between TTP and complex MND. In conclusion, the present data suggest that increased TTP is associated with suboptimal neurological development. This implies that factors associated with subfertility may play a role in the genesis of neurodevelopmental problems. Further exploration of the associations between subfertility and health outcome in offspring is necessary for the correct counselling of subfertile couples.

ACKNOWLEDGEMENTS We thank participating parents and children for their cooperation and enthusiasm.

4

FIGURE I. TTP in neurologically normal children and in children with minor neurological dysfunction. Horizontal bars indicate median values, boxes interquartile ranges, vertical lines full ranges and small circles extreme values.

Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not

67

CHAPTER 5 Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not

Pamela Schendelaar


Maas Jan Heineman


Karin J. Middelburg

Jorien Seggers

Mijna Hadders-Algra


68

ABSTRACT

Study question Does ovarian hyperstimulation, the in vitro procedures required for in vitro fertilization (IVF)/ intracytoplasmic sperm injection (ICSI) or the combination of both affect the neurological outcome of 4-year-old singletons? Summary answer Ovarian hyperstimulation, the in vitro procedure and the combination of both were not associated with worse neurological outcome in 4-year-old singletons. What is known already Assisted reproductive techniques (ARTs) are not associated with neurological dysfunction during the first postnatal years; however, effects on the long-term neurological outcome are still inconclusive. An increased time to pregnancy (TTP, a proxy for the severity of subfertility) has been associated with a less optimal neurological condition at age 2. The present study focuses on the neurodevelopmental outcome of 4-year-old ART offspring. Study design, size, duration Longitudinal, prospective follow-up study. Participants, setting, methods Four-year-old singletons born to subfertile parents (subfertile group, n=195), including singletons born after controlled ovarian hyperstimulation IVF (COH-IVF, n=63), modified natural cycle IVF (MNC-IVF, n=53) and natural conception (Sub-NC, n=79). Data on underlying cause of subfertility and TTP were present. In addition, we assessed newly recruited 4-year-old singletons born to fertile parents after natural conception (reference group, n=98). Neurological development was evaluated with the neurological examination according to Hempel, resulting in a neurological optimality score (NOS), a fluency score and the occurrence of the clinically relevant form of minor neurological dysfunction (complex MND). The primary outcome was the fluency score, as fluency of movements is easily reduced by subtle brain dysfunction. Data were analysed with univariable and multivariable regression analyses, in which special attention was paid to sex differences in the neurological outcome. Main results and the role of chance The fluency score, NOS and the prevalence of complex MND were similar in COH-IVF, MNC-IVF and Sub-NC children. The neurological condition of children born to subfertile parents was similar to that of children of fertile parents and was independent of the underlying cause of subfertility. No statistically significant associations were found between TTP and the fluency score and NOS. However, a positive correlation was found between TTP and the prevalence of complex MND (TTP in years, adjusted odds ratio [OR] [95% confidence interval, CI]: 1.207 [1.038 to 1.404], P = 0.014); a correlation which could be attributed to girls, in whom an evident positive correlation was present (adjusted OR [95%CI]: 1.542 [1.161 to 2.047], P = 0.003). A similar association was absent in boys. Limitations, reasons for caution The prospective design of our study and small postnatal attrition rate (9.3%) reduced potential selection bias based on the child’s development or health. The assessors were blind to the mode of conception, except for the group of children born to fertile parents, which was newly recruited. The study lacks sufficient power to conclude firmly that increased TTP is associated with a higher prevalence of complex MND.

5


69

Wider implications of the findings Our study suggests that the severity of subfertility, rather than its simple presence or components of IVF treatment, affects neurological outcome. Moreover, girls may be neurologically more vulnerable for the effect of severity of subfertility. The finding that the severity of subfertility may be the decisive factor rather than the presence of a history of subfertility per se corroborates previous reports. Our results cannot be generalized to multiples, as we studied singletons only.

5

70

INTRODUCTION

To date, development and health of children born following assisted reproductive techniques (ART) is of general significance. Already up to 5% of newborns in Europe are born following in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI).10 ART is known to be associated with perinatal adversities such as low birthweight and preterm delivery.52-56 Nevertheless, ART has not been associated with the adverse neurodevelopmental outcome during the first postnatal years.89 However, this does not preclude an effect of ART on long-term neurological development. Neurodevelopmental disorders may first emerge when children grow older, as a result of the protracted course of structural and functional changes of the brain during childhood.90,164,188-190

Various longitudinal studies addressed the neurodevelopmental outcome in ART children beyond the age of 2. The studies varied however considerably in the nature of ART, assessment method, age at follow-up and in study outcome. Some studies reported the worse outcome in ART children,100,113 whereas others concluded that the outcome of ART children was similar to that of naturally conceived children.78,103-109 Few studies addressed long-term neurological development. Leunens et al. and Ludwig et al. reported that neuromotor skills of ICSI singletons were similar to those of naturally conceived peers at ages varying from 5.5 to 10 years.105,107,108 However, both studies suffered from a high attrition rate. Knoester et al. who assessed minor neurological dysfunction (MND) at 5 to 8 years reported that neurological condition of children born after IVF was similar to that of naturally conceived controls.104 However, ICSI children had a higher prevalence of MND than the naturally conceived children. This means that the question whether ART affects long-term neuromotor and the neurological outcome has not been answered satisfactorily.

Several components of the ART procedure could potentially influence neurological development, for example, parental characteristics such as increased maternal age,7,57,58 ovarian hyperstimulation,171 the impact of the in vitro procedure31 and the consequences of vanishing twins.50 Moreover, underlying subfertility problems may also influence neurological development, as subfertility is known to be associated with more obstetrical and perinatal adversities.59-62 Zhu et al. found a mild psychomotor delay in 18-month-old children and a modest increased risk for developmental coordination disorder (DCD) in 7-year-old children born to subfertile parents compared to peers of fertile parents.97,98

In order to study the effects of several factors involving assisted conception on neurodevelopmental outcome, we composed the Groningen ART cohort. The cohort consists of three groups of children: a group of children born following conventional controlled ovarian hyperstimulation IVF or ICSI (COH-IVF), a group of children born following modified natural cycle IVF or ICSI (MNC-IVF) and a third group of children conceived naturally, born to subfertile parents (Sub-NC).135 Potential differences in neurodevelopmental outcome of COH-IVF and MNC-IVF children may largely be attributed to ovarian hyperstimulation, whereas potential differences in MNC-IVF and Sub-NC children may largely be attributed to the in vitro procedure. Previously, we reported that the

5


71

neurodevelopmental outcome of the three groups was similar up until 2 years of age.135,140,172 At two assessment ages we had been able to compare the developmental outcome of the Groningen ART cohort children with that of naturally conceived children born to fertile parents. The data showed that neurological condition of the Groningen ART cohort children at 3 months of age was less favourable than that of 3-month-olds of the general population. A similar difference was not observed at the age of 2 years. However, we found an association between the duration of the time to pregnancy (TTP) and a less optimal neurological condition at 2 years.191 TTP refers to the number of menstrual cycles required to conceive and is used to estimate the ability to achieve pregnancy. In other words, TTP may be used as a proxy for the severity of subfertility.

The primary aim of the study was to evaluate the effect of ovarian hyperstimulation, the in vitro procedure and the combination of both on neurological outcome at 4 years of age. Additionally, we aimed to study the effect of three other aspects of subfertility on neurological condition at 4 years. In the first place, we addressed the effect of the presence of a history of subfertility. To this end, we recruited a reference group of 4-year-olds born to fertile parents. Secondly, we evaluated the effect of the underlying cause of subfertility and thirdly, that of the duration of subfertility in terms of TTP. For the analyses of the effect of cause and duration of subfertility, we pooled the three ART study groups to form one subfertile group. Finally, we evaluated whether ICSI affected outcome by pooling the COH-IVF and MNC-IVF groups. We were primarily interested in the neurological outcome in ART offspring in general; however, specific attention was paid to the possibility of sex-specific effects of ART and subfertility, as some studies suggested that the neurodevelopmental outcome in boys conceived with ART was worse than that in girls.100,108,192

Our primary outcome was neurological condition expressed in terms of fluency of motor behaviour, using the standardized, precise and age-specific neurological assessment according to Hempel.182 The fluency score is the most sensitive measure to detect changes in neuromotor development, since reduced fluency of movements is one of the first signs associated with a non-optimal neurological condition.161,162 The additional outcome parameters were the neurological optimality score (NOS) and the rate of complex MND.

MATERIALS AND METHODS Participants Pregnant subfertile couples with a term date between March 2005 and December 2006 were recruited at the Department of Reproductive Medicine of the University Medical Center Groningen.135 All couples who achieved a singleton pregnancy following IVF or ICSI were invited to participate in the study. This resulted in a group of children born following COH-IVF or ICSI and a group of children born following MNC-IVF or ICSI, in which medication use was minimal and the follicle that developed naturally to dominance was used for assisted conception.136,137 Couples that were treated with cryopreserved or donated oocytes

5

72

or embryos to achieve pregnancy were excluded. A third group was formed by naturally conceived children born to subfertile parents (Sub-NC). These couples had tried to conceive for at least one year and eventually conceived naturally while on the waiting list for fertility evaluation or treatment.

Prenatal, perinatal and demographic information was gathered using standardized charts during the first follow-up assessment at 2 weeks post-term.135 Information on the causes and treatment of infertility was retrieved from medical records. The causes of subfertility were classified as 'tubal pathology' in case of abnormalities of the fallopian tubes, as ‘male factor’ in case of male infertility, as ‘other causes’ in case of endometriosis, cervical factor or hormonal cause and as ‘unknown cause’ in case of lack of a specific cause for subfertility. TTP was defined as the interval between the start of timed unprotected intercourse or a previous pregnancy and conception, recorded in years and months and finally converted into decimal years. Note that in case of miscarriage, a new onset of TTP was defined. This explains why some participating subfertile parents in the present study had a TTP of < 1 year (Table I).

For the present study, a new retrospective reference group was recruited between December 2009 and February 2012 at six child welfare centres in and around Groningen. All parents of 4-year-old children who visited the child welfare clinic for routine general health care were invited to participate. Children of parents who had tried to achieve a pregnancy for > 1 year or who achieved a pregnancy by any form of assisted conception were excluded.

Neurological assessment All children were assessed with the standardized and age-specific neurological examination according to Hempel,182 around the time of their fourth birthday. The Hempel examination assesses MND at preschool age by means of five domains of functions: fine motor function, gross motor function, posture and muscle tone, reflexes and visuomotor function.155 Each of the domains can be scored as typical or deviant.

Children were classified as being neurologically normal, having simple MND, having complex MND or being neurologically abnormal. Neurologically normal implies the absence of neurological dysfunction; it implies the absence of deviant domains or the isolated presence of dysfunction in the domain of reflexes. Simple MND indicates the presence of one deviant domain (except the domain of reflexes) and is regarded as a non-optimal but normal form of brain function.90 Complex MND indicates the presence of more than one domain of dysfunction. It represents the clinically relevant form of MND, as it is associated with prematurity and other perinatal adversities, and learning and behavioural disorders at later age.90,156-159,193 Neurologically abnormal implies the presence of a distinct neurological syndrome such as cerebral palsy.

The outcome of the Hempel assessment was also expressed in a NOS. The NOS consists of 56 items (range 0–56), for which an optimal condition is defined. The total score results from the sum of the items which fulfil the criteria for optimality. Higher scores represent better performance. Note that the range for optimal behaviour is narrower than that for

5


73

normal behaviour.160 The application of the optimality concept turns the NOS into a sensitive tool to evaluate neurological integrity. The fluency score (range 0-15) is a subscore of the NOS that evaluates the fluency of motor behaviour. The inter-rater reliability of the Hempel assessment is satisfactory (κ = 0.62-1.00 [mean 0.93]) and its construct validity is good.155,163-165

The assessment at 4 years was carried out by trained assessors supervised by a neurodevelopmental expert (M.H.-A.). The assessors and supervisor were blind to prenatal and perinatal history, including the mode of conception of the three subfertile groups. It was not possible to blind the assessors for the fertile reference group, as this group was recruited separately from the three subfertile study groups. The assessments were carried out between February 2009 and February 2012 at the Institute of Developmental Neurology at the University Medical Center Groningen, The Netherlands. Statistical analysis Power calculation was based on the neurological outcome at 18 months. To detect at least half a standard deviation difference in the fluency score (mean 9.5, SD 1.7) with 80% power, at least 64 children had to be included in each group.140,161

Fisher’s exact tests, Mann-Whitney U-tests or Student’s t-tests were used to test differences between groups. Potential associations between the neurological outcome and the specific contrasts (various ART groups; subfertile versus fertile reference group; underlying subfertility causes; TTP) were analysed using regression analyses: linear regression analyses were applied for the fluency score and NOS; logistic regression analyses were applied for the prevalence of complex MND. Data were analysed for the total group of children; to study a potential effect for boys and girls separately, an interaction effect with sex was incorporated in the regression analysis. In addition to the univariable regression analyses, we applied multivariable regression analyses in which we adjusted for confounders on an a priori basis. We adjusted for the variables ‘gestational age’, ‘birthweight’, ‘vanishing twins’, ‘TTP’, ‘parental age’ and ‘high parental educational level’ in the analyses on the pairwise comparisons between the three ART study groups. In the analyses on the difference between the subfertile and the fertile reference group, we adjusted for the same set of confounders, except for the variable TTP. Additionally, we adjusted for the variables ‘Caesarean section’, ‘breastfed for > 6 weeks’ and ‘firstborn’. The latter three variables were added as they differed significantly between the subfertile and fertile reference group. In the analyses on the effect of specific underlying causes of subfertility and the effect of TTP on the neurological outcome we adjusted for ‘gestational age’, ‘birthweight’, ‘vanishing twins’, ‘parental age’ and ‘high parental educational level’. Note that in the analyses on underlying subfertility causes and TTP we used only the data of the subfertile group, i.e. we pooled the COH-IVF, MNC-IVF and Sub-NC groups. Lastly, Fisher’s exact tests and Mann-Whitney U-tests were used to evaluate whether ICSI affected the outcome. To this end the COH-IVF and MNC-IVF groups were pooled.

5

74

All analyses were performed using the IBM Statistical Package for the Social Sciences (SPSS), version 20 and the Statistical Analysis System (SAS) software, version 9.3. In the analyses P < 0.05 were considered significant. Bonferroni corrections were applied in the analyses on pairwise comparisons of the three ART study groups and the analyses of the underlying subfertility causes: P < 0.017 (0.05/3) and 0.013 (0.05/4) respectively, were considered significant.

Ethical approval The Medical Ethical Commission of the University Medical Center Groningen approved the study design. Parents provided written informed consent for participation of their child in the study.

RESULTS

The three study groups of the Groningen ART cohort Participation and demographic characteristics During the prenatal period, 89 COH-IVF children, 79 MNC-IVF children and 143 Sub-NC children were eligible to participate in the study. Of these, respectively 68 (76%), 57 (72%) and 90 (93%) children were included in the follow-up study (Figure I).135 Generally, social, obstetrical and neonatal characteristics of participants and non-participants were similar. However, in the Sub-NC group, non-participating mothers were younger than participating mothers (P = 0.030).135

At the follow-up assessment at 4 years, 5 (7%) COH-IVF children, 4 (7%) MNC-IVF children and 11 (12%) Sub-NC children were lost to follow-up due to logistical problems or assessment burden (total postnatal attrition rate 9.3%, Figure I). One MNC-IVF girl died at 3 weeks of age as a consequence of a congenital heart disorder. Parent and child characteristics for participants and non-participants within the groups were similar, except for TTP in the MNC-IVF group: participating MNC-IVF parents had a longer TTP than non-participating MNC-IVF parents (median [range]: 3.8 [0.1 to 13.2] and 1.3 [0.7 to 2.0] respectively, P = 0.021).

Demographic characteristics of parents and children of the three ART groups are listed in Table I. Overall, demographic characteristics of the three groups were similar, except for the following: COH-IVF children had shorter gestational age at birth (P = 0.017) than Sub-NC children. COH-IVF and MNC-IVF children had lower birthweight than Sub-NC children (P = 0.044 and P = 0.048, respectively). COH-IVF children were more often survivors of a vanishing twin than MNC-IVF children and Sub-NC children (P = 0.038 and P = 0.001, respectively). TTP was longer in the COH-IVF and MNC-IVF groups than in the Sub-NC group (P < 0.001 and P = 0.001, respectively). COH-IVF fathers were older than MNC-IVF fathers at time of conception (P = 0.042).

5


75

COH - IVF MNC - IVF Sub - NCn = 63 n = 53 n = 79

Child characteristicsMale gender, n (%) 34 (54.0) 26 (49.1) 41 (51.9)First born, n (%) 43 (68.3) 37 (69.8) 49 (62.0)Corrected age at examination at 4 years of age (months), median (range) 50 (47.5 - 60.1) 48.9 (48.0 - 52.5) 48.9 (47.9 - 56.4)

Birth characteristicsGestational age (weeks), median (range) 39.4 (33.4 - 42.3)* 40.1 (34.6 - 42.6) 40.0 (30.1 - 42.7)*

Preterm birth (< 37 weeks), n (%) 7 (11.1) 6 (11.3) 5 (6.3)Birthweight (grams), mean (SD) 3393.1 (563.2)* 3384.4 (585.7)* 3577.9 (519.4)*/*

Low birthweight (< 2500 gram), n (%) 3 (4.8) 4 (7.5) 3 (3.8)Small for gestational agea, n (%) 0 3 (5.7) 1 (1.3)Caesarean section, n (%) 15 (23.8) 8 (15.1) 21 (26.6)Signs of fetal distressb, n (%) 19 (30.2) 15 (28.3) 34 (43.0)

Neonatal characteristicsApgar score 5 min < 7c, n (%) 0 0 1 (1.3)Neonatal intensive care admission, n (%) 1 (1.6) 2 (3.8) 5 (6.3)Breastfed for > 6 weeksc, n (%) 28 (46.7) 24 (45.3) 39 (50.0)

Parental characteristicsMaternal age at conception in years, median (range) 32.3 (26.3 - 40.9) 32.8 (25.3 - 37.5) 33.0 (22.2 - 40.3)Paternal age at conception in yearsc, median (range) 35.4 (27.5 - 56.1)* 34.0 (28.3 - 47.8)* 35.0 (25.5 - 48.7)Education level mother (highd), n (%) 20 (31.7) 20 (37.7) 37 (46.8)Education level father (highd)e, n (%) 28 (46.7) 17 (32.7) 29 (36.7)Smoking during pregnancy, n (%) 7 (11.1) 7 (13.2) 9 (11.4)Alcohol consumption during pregnancy, n (%) 3 (4.8) 0 (0) 2 (2.5)

Fertility parametersTime to pregnancy in yearsc,e,median (range) 4.0 (0.1 - 13.3)*** 3.8 (0.1 - 13.2)** 2.1 (0.1 - 11.3)***/**

Primary subfertility, n (%) 35 (55.6) 32 (60.4) 41 (51.9)ICSI performed n (%) 41 (65.1) 26 (49.1) n.a.

Vanishing twins, n (%) 8 (12.7)*/** 1 (1.9)* 0**

Subfertility causesf

Tubal pathology, n (%) 12 (19.0) 10 (18.9) 6 (7.6)Male factor, n (%) 30 (47.6)** 29 (54.7)*** 16 (20.3)**/***

Other causes, n (%) 19 (30.2)* 6 (11.3)* 15 (19.0)Unknown cause, n (%) 9 (14.3)*** 10 (18.9)*** 47 (59.5)***/***

* P < 0.05; ** P < 0.01; *** P < 0.001.

d University education or vocational colleges.d Time to pregnancy of the three ART study groups was recorded in years and months and finally converted into decimal years. In case of a miscarriage the onset of time to pregnancy restarted, therefore time to pregnancy may be shorter than one year.

Characteristics

c Missing data in three groups: Apgar score 5 min < 7 n=3, breastfed for > 6 weeks n=4, paternal age at conception n=3, education level father n=4, time to pregnancy n=1.

f Couples may have more than one cause of subfertility, therefore totals may exceed 100% .

TABLE I. Characteristics of participating parents and children of the ART study groups.

Fisher’s exact tests, Mann-Whitney U -tests or Student’s t -tests were performed to investigate differences between groups. ART: assisted reproductive techniques, COH-IVF: children born following controlled ovarian hyperstimulation IVF or ICSI, MNC-IVF: children born following modified natural cycle IVF or ICSI, and Sub-NC: naturally conceived children born to subfertile parents.a Birthweight for gestational age is below 2 standard deviation scores compared with a Dutch reference population (Dutch reference tables, perinatal Registration Netherlands).b Signs of fetal distress denoted by meconium stained amniotic fluid and/or cardiotocographic signs and/or acidosis.

5

76

The neurological outcome in the three ART study groups Note that we first present the neurological outcome in clinical terms of complex MND. Thereafter, we describe the results of the primary outcome parameter, the fluency score, and subsequently the NOS.

At 4 years, none of the children of the COH-IVF, MNC-IVF and Sub-NC groups were classified as neurologically abnormal. Twenty-one (33%) COH-IVF children, 8 (15%) MNC-IVF children and 18 (23%) Sub-NC children had complex MND (Tables II and III). The differences in the prevalence of complex MND between the three ART study groups did not reach statistical significance, without and with adjustment for confounders (Table IV). The absence of a group effect held also true for the subgroups of boys and girls (Tables II and IV). Neither statistically significant interaction effects were found between sex and ART group status (COH-IVF, MNC-IVF or Sub-NC group; Table IV). The mean values of the primary outcome parameter, the fluency score, were 12.0, 12.4 and 12.1 for COH-IVF, MNC-IVF and Sub-NC groups, respectively (Table II). The mean fluency scores did not significantly differ between the three groups without and with adjustment for confounders (Table IV). The same held true for boys and girls separately (Tables II and

5

FIGURE I. Flow chart of the Groningen ART cohort.


77

IV). No significant interaction effects were found between sex and ART group status (Table IV). Similar results were found for the NOS (Tables II and IV).

Subfertile group: the effect of underlying subfertility causes on the neurological outcome None of the underlying subfertility causes such as tubal pathology, male factor, other causes and unknown cause, were associated with the fluency score, NOS or the prevalence of complex MND (data not presented). No interactions between sex and underlying cause of subfertility were found for the neurological outcome parameters, except for sex and tubal pathology for the prevalence of complex MND (P = 0.038).

Subfertile group: the effect of ICSI on the neurological outcome No statistically significant differences were found between children born after IVF with and IVF without ICSI in fluency score, the NOS and the occurrence of complex MND (P = 0.250, P = 0.968 and P = 1, respectively). Median TTP’s were also similar (IVF with ICSI: 4.1 [range 0.1 to 13.3], IVF without ICSI: 3.8 [range 1 to 13.2], P = 0.261).

Subfertile group: the effect of time to pregnancy on neurological outcome TTP (expressed in decimal years) was positively correlated to the prevalence of complex MND without and with adjustment for confounders (adjusted odds ratio [OR] [95% confidence interval, CI]: 1.207 [1.038 to 1.404], P = 0.014, Table III). A similar statistically significant positive correlation between TTP and prevalence of complex MND was found for the subgroup of girls, but not for the boys: girls had a higher prevalence of complex MND when their parents had a longer TTP, while a similar association was absent in boys (Table IV). The interaction of sex and TTP was statistically significant (P = 0.034, Table IV). Table IV illustrates the association between duration of TTP and complex MND. It indicates that the most clear increase in the prevalence of complex MND was found after a parental TTP of at least 3 years: 32.0% (n=33) versus 15.7% (n=14) for children born after a TTP of just 1 - 3 years (adjusted OR [95%CI]: 2.853 [1.306 to 6.235], P = 0.009). This association was present in the subgroup of girls (adjusted OR [95%CI]: 7.319 [2.072 to 25.855], P = 0.002), but not in that of the boys (adjusted OR [95%CI]: 1.194 [0.387 to 3.681], P = 0.757).

TTP was not associated with the fluency score and the NOS, neither in the total group of children, nor in boys and girls separately. The fluency score and the NOS were also not affected by a statistically significant interaction between sex and TTP (Table IV).

5

78

boys

girls

total

boys

girls

total

n = 10

1n =

94n =

195

n = 54

n = 44

n = 98

Neur

olog

ical

opt

imal

ityFlu

ency

scor

e, me

an [9

5% C

I]11

.8 [11

.5 ; 1

2.1]

12.5

[12.2

; 12.8

] 12

.2 [11

.9 ; 1

2.4]

12.5

[12.1

; 12.9

]13

.1 [12

.7 ; 1

3.6]

12.8

[12.5

; 13.1

]

Neur

ologic

al op

timali

ty sc

ore,

mean

[95%

CI]

47.5

[46.9

; 48.2

] 48

.8 [48

.1 ; 4

9.5]

48.2

[47.7

; 48.6

] 48

.6 [47

.6 ; 4

9.5]

50.2

[49.2

; 51.2

] 49

.4 [48

.7 ; 5

0.1]

Neur

olog

ical

cla

ssifi

catio

n Neur

ologic

ally n

orma

l, n (%

)60

(59.4

)51

(54.3

)11

1 (56

.9)32

(59.3

)34

(77.3

)66

(67.3

)

Simple

MND

, n (%

)20

(19.8

)17

(18.1

)37

(19.0

)12

(22.2

)23

(16.7

)18

(18.4

)

Comp

lex M

ND, n

(%)

21 (2

0.8)

26 (2

7.7)

47 (2

4.1)

10 (1

8.5)

4 (9.1

)14

(14.3

)

Neur

ologic

ally a

bnor

mal, n

(%)

00

00

00

Dom

ains

of n

euro

logi

cal d

ysfu

nctio

nFin

e moto

r dys

functi

on, n

(%)

00

0 (0)

00

0 (0)

Gros

s moto

r dys

functi

on, n

(%)

30 (2

9.7)

24 (2

5.5)

54 (2

7.7)

16 (2

9.6)

4 (9.1

)20

(20.4

)

Postu

re an

d mus

cle to

ne dy

sfunc

tion,

n (%

)22

(21.8

)25

(26.6

)47

(24.1

)10

(18.5

)8 (

18.2)

18 (1

8.4)

Dysfu

nctio

nal re

flexe

s, n (

%)

45 (4

4.6)

36 (3

8.3)

81 (4

1.5)

22 (4

0.7)

16 (3

6.4)

38 (3

8.8)

Visuo

motor

dysfu

nctio

n, n (

%)

00

00

00

b Refe

renc

e gro

up: n

atura

lly co

nceiv

ed ch

ildre

n bor

n to f

ertile

pare

nts.

Subf

ertile

gro

upa

Refer

ence

gro

upb

TABL

E III.

Neu

rolo

gica

l opt

imali

ty, n

euro

logi

cal c

lassif

icatio

n an

d do

main

s of n

euro

logi

cal d

ysfu

nctio

n of

the s

ubfe

rtile

and

refe

renc

e gro

ups.

ART:

assis

ted re

prod

uctiv

e tec

hniqu

es, C

OH-IV

F: ch

ildre

n bor

n foll

owing

contr

olled

ovar

ian hy

perst

imula

tion I

VF or

ICSI

, MNC

-IVF:

child

ren b

orn f

ollow

ing m

odifie

d natu

ral c

ycle

IVF

or IC

SI, a

nd S

ub-N

C: na

turall

y con

ceive

d chil

dren

born

to su

bfertil

e par

ents.

NOS

= ne

urolo

gical

optim

ality

scor

e, MN

D =

mino

r neu

rolog

ical d

ysfun

ction

.a S

ubfer

tile gr

oup:

COH-

IVF,

MNC

-IVF

and S

ub-N

C gr

oups

toge

ther.

boys

girls

total

boys

girls

total

boys

girls

total

n = 34

n = 29

n = 63

n = 26

n = 27

n = 53

n = 41

n = 38

n = 79

Neur

olog

ical

opt

imal

ityFlu

ency

scor

e, me

an [9

5% C

I]11

.7 [11

.2 ; 1

2.2]

12.3

[11.7

; 12.8

] 12

.0 [11

.6 ; 1

2.4]

12.0

[11.5

; 12.6

] 12

.7 [12

.2 ; 1

3.3]

12.4

[12.0

; 12.8

] 11

.8 [11

.3 ; 1

2.2]

12.5

[12.0

; 12.9

] 12

.1 [11

.8 ; 1

2.4]

Neur

ologic

al op

timali

ty sc

ore,

mean

[95%

CI]

47.0

[45.8

; 48.2

] 48

.1 [46

.8 ; 4

9.4]

47.6

[46.7

; 48.4

] 48

.2 [46

.8 ; 4

9.5]

49.0

[47.7

; 50.3

] 48

.6 [47

.6 ; 4

9.5]

47.6

[46.5

; 48.6

] 49

.1 [48

.0 ; 5

0.3]

48.4

[47.4

; 49.2

]

Neur

olog

ical

cla

ssifi

catio

nNe

urolo

gicall

y nor

mal, n

(%)

18 (5

2.9)

15 (5

1.7)

33 (5

2.4)

19 (7

3.1)

15 (5

5.6)

34 (6

4.2)

23 (5

6.1)

21 (5

5.3)

44 (5

5.7)

Simple

MND

, n (%

)5 (

14.7)

4 (13

.8)9 (

14.3)

4 (15

.4)7 (

25.9)

11 (2

0.8)

11 (2

6.8)

6 (15

.8)17

(21.5

)

Comp

lex M

ND, n

(%)

11 (3

2.4)

10 (3

4.5)

21 (3

3.3)

3 (11

.5)5 (

18.5)

8 (15

.1)7 (

17.1)

11 (2

8.9)

18 (2

2.8)

Neur

ologic

ally a

bnor

mal, n

(%)

00

00

00

00

0

Dom

ains

of n

euro

logi

cal d

ysfu

nctio

nFin

e moto

r dys

functi

on, n

(%)

00

00

00

00

0

Gros

s moto

r dys

functi

on, n

(%)

11 (3

2.4)

6 (20

.7)17

(27.0

)5 (

19.2)

7 (25

.9)12

(22.6

)14

(34.1

)11

(28.9

)25

(31.6

)

Postu

re an

d mus

cle to

ne dy

sfunc

tion,

n (%

)9 (

26.5)

10 (3

4.5)

19 (3

0.2)

3 (11

.5)6 (

22.2)

9 (17

.0)10

(24.4

)9 (

23.7)

19 (2

4.1)

Dysfu

nctio

nal re

flexe

s, n (

%)

17 (5

0.0)

13 (4

4.8)

25 (3

1.6)

13 (5

0.0)

9 (33

.3)19

(24.1

)15

(36.6

)14

(36.8

)29

(36.7

)Vis

uomo

tor dy

sfunc

tion,

n (%

)0

00

00

00

00

TABL

E II.

Neur

olog

ical o

ptim

ality

, neu

rolo

gica

l clas

sifica

tion

and

dom

ains o

f neu

rolo

gica

l dys

func

tion

of th

e ART

stud

y gro

ups.

ART:

assis

ted re

prod

uctiv

e tec

hniqu

es, C

OH-IV

F: ch

ildre

n bor

n foll

owing

contr

olled

ovar

ian hy

perst

imula

tion I

VF or

ICSI

, MNC

-IVF:

child

ren b

orn f

ollow

ing m

odifie

d natu

ral c

ycle

IVF

or IC

SI, a

nd S

ub-N

C: na

turall

y con

ceive

d chil

dren

born

to su

bfertil

e par

ents.

NOS

= ne

urolo

gical

optim

ality

scor

e, MN

D =

mino

r neu

rolog

ical d

ysfun

ction

.

COH-

IVF

MNC-

IVF

Sub-

NC

5


79

BOYS

GIRL

STO

TAL

inter

actio

n eff

ect w

ith se

x

Adju

sted

mea

n di

ffere

nce

[95%

CI

]P

-val

ueAd

just

ed m

ean

diffe

renc

e [9

5%

CI]

P-v

alue

Adju

sted

mea

n di

ffere

nce

[95%

CI

]P

-val

ueP

-val

ue

COH-

IVF

versu

s Sub

-NCa

-0.09

3 [-0

.881 ;

0.69

5]0.8

15

0.179

[-0.7

19 ;

1.077

]0.6

92-0

.042 [

-0.61

6 ; 0.

532]

0.885

0.963

MNC-

IVF

versu

s Sub

-NCa

0.034

[-0.7

38 ;

0.805

]0.9

310.3

22 [-

0.501

; 1.1

45]

0.439

0.216

[-0.3

42 ;

0.773

]0.4

47CO

H-IV

F ve

rsus M

NC-IV

Fa-0

.127 [

-0.95

4 ; 0.

701]

0.762

-0.14

3 [-1

.064 ;

0.77

9]0.7

59-0

.258 [

-0.86

1 ; 0.

346]

0.400

Subfe

rtile g

roup

b ve

rsus R

efere

nce g

roup

c-0

.389 [

-1.01

5 ; 0.

237]

0.222

-0.30

6 [-0

.937 ;

0.32

5]0.3

39-0

.383 [

-0.82

7 ; 0.

061]

0.090

0.879

Mea

n es

timat

e [9

5% C

I]P

-val

ue M

ean

estim

ate

[95%

CI]

P-v

alue

Mea

n es

timat

e [9

5% C

I]P

-val

ueP

-val

ue0.1

10 [-

0.019

; 0.2

39]

0.095

0.024

[-0.1

27 ;

0.176

]0.7

520.0

81 [-

0.019

; 0.1

80]

0.110

0.347

Adju

sted

mea

n di

ffere

nce

[95%

CI

]P

-val

ueAd

just

ed m

ean

diffe

renc

e [9

5%

CI]

P-v

alue

Adju

sted

mea

n di

ffere

nce

[95%

CI

]P

-val

ueP

-val

ue

COH-

IVF

versu

s Sub

-NCa

-0.50

9 [-2

.456 ;

1.43

7]0.6

04-0

.415 [

-2.47

2 ; 1.

642]

0.689

-0.54

7 [-1

.910 ;

0.81

6]0.4

290.8

19MN

C-IV

F ve

rsus S

ub-N

Ca0.4

00 [-

1.504

; 2.3

04]

0.677

0.077

[-1.8

10 ;

1.963

]0.9

360.3

40 [-

0.985

; 1.6

64]

0.614

COH-

IVF

versu

s MNC

-IVFa

-0.90

9 [-2

.953 ;

1.13

4]0.3

79-0

.492 [

-2.60

3 ; 1.

619]

0.644

-0.88

7 [-2

.319 ;

0.54

6]0.2

23Su

bfertil

e gro

upb

versu

s Refe

renc

e gro

upc

-0.72

0 [-2

.251 ;

0.81

2]0.3

54-0

.963 [

-2.36

3 ; 0.

438]

0.176

-0.87

7 [-0

.914 ;

0.16

0]0.0

970.6

83 M

ean

estim

ate

[95%

CI]

P-v

alue

Mea

n es

timat

e [9

5% C

I]P

-val

ue M

ean

estim

ate

[95%

CI]

P-v

alue

P-v

alue

0.102

[-0.2

18 ;

0.422

]0.5

28-0

.240 [

-0.58

7 ; 0.

106]

0.172

-0.03

9 [-0

.276 ;

0.19

7]0.7

440.2

22

Adju

sted

odd

s ra

tio [9

5% C

I]P

-val

ueAd

just

ed o

dds

ratio

[95%

CI]

P-v

alue

Adju

sted

odd

s ra

tio [9

5% C

I]P

-val

ueP

-val

ue

COH-

IVF

versu

s Sub

-NCa

1.798

[0.91

4 ; 3.

536]

0.089

0.675

[0.31

1 ; 1.

467]

0.321

1.132

[0.73

4 ; 1.

743]

0.575

MNC-

IVF

versu

s Sub

-NCa

0.431

[0.06

7 ; 2.

761]

0.374

0.199

[0.03

7 ; 1.

061]

0.059

0.349

[0.12

0 ; 1.

018]

0.054

COH-

IVF

versu

s MNC

-IVFa

6.217

[1.05

2 ; 2.

930]

0.044

2.824

[0.52

7 ; 15

.131]

0.225

2.934

[1.02

8 ; 8.

370]

0.044

Subfe

rtile g

roup

b ve

rsus R

efere

nce g

roup

c0.9

46 [0

.305 ;

2.93

0]0.9

233.1

39 [0

.920 ;

10.82

4]0.0

701.5

24 [0

.699 ;

3.32

5]0.2

890.1

321.0

00 [0

.801 ;

1.24

7]0.9

981.5

42 [1

.161 ;

2.04

7]0.0

031.2

07 [1

.038 ;

1.40

4]0.

014

0.034

d Only

time t

o pre

gnan

cy da

ta of

coup

les of

the s

ubfer

tile gr

oup w

ere u

sed f

or th

e ana

lyses

.

Note

that in

gene

ral P

< 0.

05 w

ere c

onsid

ered

to be

sign

ifican

t. Bon

ferro

ni co

rrecti

ons w

ere a

pplie

d in t

he an

alyse

s on p

airwi

se co

mpar

isons

of th

e thr

ee G

ronin

gen A

RT co

hort

study

grou

ps, P

< 0.

05/3

= 0.0

17 w

ere c

onsid

ered

to be

sign

ifican

t.

Time t

o pre

ganc

yd (yea

rs)

Time t

o pre

gnan

cyd (y

ears)

e

TABL

E IV

. Out

com

es o

f the

effe

ct o

f asp

ects

of a

ssist

ed co

ncep

tion

and

the p

rese

nce a

nd se

verit

y of s

ubfe

rtilit

y on

the n

euro

logi

cal o

utco

me o

f 4-y

ear-o

lds.

Outc

ome

mea

sure

: FLU

ENCY

SCO

RE

Outc

ome

mea

sure

: NOS

Outc

ome

mea

sure

: com

plex

MND

a Adju

sted f

or ge

statio

nal a

ge, b

irthwe

ight, v

anish

ing tw

ins, ti

me to

preg

nanc

y, pa

renta

l age

and p

aren

tal ed

ucati

onal

level.

b Sub

fertile

grou

p: CO

H-IV

F, M

NC-IV

F an

d Sub

-NC

grou

ps to

gethe

r.c A

djuste

d for

gesta

tiona

l age

, birth

weigh

t, van

ishing

twins

, Cae

sare

an se

ction

, bre

astfe

d for

> 6

week

s, firs

tborn

, par

ental

age a

nd pa

renta

l edu

catio

nal le

vel.

0.232

e Adju

sted f

or ge

statio

nal a

ge, b

irthwe

ight, v

anish

ing tw

ins, p

aren

tal ag

e and

pare

ntal e

duca

tiona

l leve

l.

Time t

o pre

gnan

cyd (y

ears)

e

Time t

o pre

gnan

cyd (y

ears)

e

COH-

IVF:

child

ren b

orn f

ollow

ing co

ntroll

ed ov

arian

hype

rstim

ulatio

n IVF

or IC

SI, M

NC-IV

F: ch

ildre

n bor

n foll

owing

mod

ified n

atura

l cyc

le IV

F or

ICSI

, and

Sub

-NC:

natur

ally c

once

ived c

hildr

en bo

rn to

subfe

rtile p

aren

ts. N

OS =

neur

ologic

al op

timali

ty sc

ore,

MND

= mi

nor n

euro

logica

l dys

functi

on.

5

80

Subfertile group and fertile reference group Participation and demographic characteristics Parents of 222 reference children (all singletons) were invited to participate in the follow-up study. Of the invited parents, 7 children had to be excluded as it had taken > 1 year to achieve pregnancy. Of the 215 eligible children, parents of 117 (54%) children refused participation; eventually 98 (46%) parents allowed their child to participate in the fertile reference group. Demographic characteristics such as sex, gestational age, firstborn, maternal age and parental educational levels were similar for participants and non-participants (data not presented). Demographic characteristics of parents and children of the subfertile group and the fertile reference group are listed in Table VI. Children of the subfertile group were more often firstborn (P = 0.002), had a slightly lower birthweight (P = 0.049), were more often born following Caesarean section (P = 0.006) and were less often breastfed for > 6 weeks (P = 0.006) than children of the fertile reference group. Parents of children born in the subfertile groups were older (maternal and paternal age, both P < 0.001) and less often highly educated (education level mother: P = 0.034; education level father: P = 0.005) than parents of children born in the fertile reference group. The neurological outcome in the subfertile group and reference group. At 4 years, none of the children of the subfertile group or reference group were classified as neurologically abnormal. Fourty-seven (24%) children of the subfertile group and 14 (14%) children of the fertile reference group had complex MND (Table III), a difference that did not reach statistical significance without and with adjustment for confounders (adjusted OR [95%CI]: 1.524 [0.699 to 3.325], P = 0.289, Table IV). Interestingly, the trend towards a more unfavourable neurological condition in the subfertile group was only present in girls, but not in boys (girls subfertile group: 26 [28%], girls fertile reference group: 4 [9%]; OR [95%CI]: 3.824 [1.244 to 11.751], P = 0.019). However, the difference disappeared after adjusting for

Total groupa Boysa Girls Total group Boys Girlsn = 147 n = 79 n = 68 n = 47 n = 21 n = 26

0 to 0.9 year TTPb 16 (10.9%) 8 (10.1%) 8 (11.8%) 1 (2.1%) 0 1 (3.9%)1 to 1.9 years TTP 37 (25.2%) 21 (26.6%) 16 (23.5%) 8 (17.0%) 6 (28.6%) 2 (7.7%)2 to 2.9 years TTP 24 (16.3%) 11 (13.9%) 13 (19.1%) 5 (10.6%) 3 (14.3%) 2 (7.7%)3 to 4.9 years TTP 41 (27.9%) 21 (26.6%) 20 (29.4%) 19 (40.4%) 8 (38.1%) 11 (42.3%)> 5 years TTP 29 (19.7%) 18 (22.8%) 11 (16.2%) 14 (29.8%) 4 (19.1%) 10 (38.5%)

b TTP was defined as the interval between the start of timed unprotected intercourse or a previous pregnancy and conception, recorded in years and months and finally converted into decimal years. Note that in case of miscarriage, a new onset of TTP was defined. This explains why some participating parents had a TTP of < 1 year.

MND: minor neurological dysfunction, TTP: time to pregnancy

TABLE V. Prevalence of complex MND per parental TTP category.

No complex MND Complex MND

Only TTP data of couples of the subfertile group (COH-IVF, MNC-IVF and Sub-NC groups) were used for the analyses.a TTP data were missing for one boy (no complex MND).

5


81

confounders (adjusted OR [95%CI]: 3.139 [0.920 to 10.824], P = 0.070, Table III). No statistically significant interaction effects were found between sex and group status (subfertile vs. fertile reference group, Table IV).

Univariable analysis indicated that the mean fluency score of the subfertile group was significantly lower than that of the reference group (12.2 versus 12.8, resp.; mean difference [95% CI]: -0.617 [-0.986 to -0.247], P = 0.001; Table III). However, after adjusting for confounders, the difference between the two groups was no longer statistically significant (adjusted median difference [95%CI]: -0.383 [-0.827 to 0.061], P = 0.090, Table IV). Similar results were found for boys and girls separately: the difference in fluency scores between ‘subfertile’ and reference boys was statistically significant (mean difference [95% CI]: boys: -

Subfertile group Reference groupn = 195 n = 98

Child characteristicsMale gender, n (%) 101 (51.8) 54 (55.1) 0.621First born, n (%) 129 (66.2) 46 (46.9) 0.002Corrected age at examination at 4 years of age (months), median (range) 48.9 (47.5 - 60.1) 49.1 (48.0 - 54.6) 0.217

Birth characteristicsGestational age (weeks)a, median (range) 40.0 (30.1 - 42.7) 40.1 (32.0 - 42.4) 0.432Preterm birth (< 37 weeks)a, n (%) 18 (9.2) 3 (3.1) 0.058Birthweight (grams)a, mean (sd) 3465.6 (557.2) 3599.7 (507.0) 0.049Low birthweight (< 2500 gram)a, n (%) 10 (5.1) 2 (2.1) 0.348Small for gestational agea,b, n (%) 4 (2.1) 4 (4.3) 0.281Caesarean section, n (%) 44 (22.6) 9 (9.2) 0.006Signs of fetal distressc, n (%) 68 (34.9) 29 (29.6) 0.430

Neonatal characteristicsApgar score 5 min < 7a, n (%) 1 (0.5) 3 (4.2) 0.062Neonatal intensive care admission, n (%) 8 (4.1) 10 (10.2) 0.067Breastfed for > 6 weeksa, n (%) 91 (47.6) 64 (65.3) 0.006

Parental characteristicsMaternal age at conception in yearsa, median (range)

32.8 (22.2 - 40.9) 30.4 (18.8 - 40.5) <0.001

Paternal age at conception in yearsa, median (range)

35 (25.5 - 56.1) 32.6 (22.5 - 45.1) <0.001

Education level mother (highd), n (%) 77 (39.5) 52 (53.1) 0.034Education level father (highd)a, n (%) 74 (38.7) 53 (56.4) 0.005Smoking during pregnancy, n (%) 23 (11.8) 5 (5.1) 0.091Alcohol consumption during pregnancy, n (%) 5 (2.6) 4 (4.1) 0.489

Fertility parametersTime to pregnancy (categoricale)a, median (range) 4 (0 - 5) 0 (0 - 1) <0.001

b Signs of fetal distress denoted by meconium stained amniotic fluid and/or cardiotocographic signs and/or acidosis.c University education or vocational colleges.

a Birth weight for gestational age is less than -2 standard deviations compared with the Dutch reference population (Dutch reference tables, perinatal Registration Netherlands).

d Time to pregnancy of the reference group was recorded in half-years, so data for the subfertile and reference group were transformed to categorical values: 0 = 0-0.5 years, 1 = 0.5-1 years, 2 = 1-2 years, 3 = 2-3 years, 4 = 3-5 years and 5 = >5 years.

Characteristics

TABLE VI. Characteristics of parents and children of the subfertile group and the reference group.

a Missing data in two groups: gestational age n=1, preterm n=1, birthweight n=4, low birthweight n=3, small for gestational age n=1, Apgar score 5 min < 7 n=31, breastfed > 6 weeks n=4, maternal age at conception n=1, paternal age at conception n=8, education level father n=8, time to pregnancy n=1.

P -value

Subfertile group: COH-IVF, MNC-IVF and Sub-NC groups together.Fisher’s exact tests, Mann-Whitney U -tests or Student’s t -tests were performed to investigate differences between groups.

5

82

0.641 [-1.138 to -0.145], P = 0.012) and that between ‘subfertile’ and reference girls as well (mean difference [95% CI]: -0.635 [-1.170 to 0.100], P = 0.020). However, the differences disappeared after adjusting for confounders (Table IV). The fluency score was not affected by a significant interaction effect between sex and the presence of subfertility (Table IV).

The mean NOS of the subfertile group was significantly lower than that of the reference group (48.2 and 49.4 respectively, mean difference [95%CI]: -1.163 [-2.008 to -0.317], P = 0.007; Tables III). Also this difference disappeared after adjusting for confounders (adjusted mean difference [95%CI]: -0.877 [-0.914 to 0.160], P = 0.097, Table IV). A similar trend was found in the subgroup of girls (mean difference [95% CI]: -1.428 [-2.609 to 0.247], P = 0.018), but the association disappeared after adjusting for confounders (Table IV). A similar trend was absent in boys. The NOS was not affected by a statistically significant interaction between sex and presence of subfertility (Table IV).

DISCUSSION The present study indicates that neurological condition at 4 years of COH-IVF, MNC-IVF and Sub-NC children does not differ. This holds true for boys and girls. Therefore, our findings suggest that neither the ovarian hyperstimulation, nor the in vitro procedures or a combination of both factors is associated with the worse neurological outcome in children up until 4 years of age. The findings correspond to the earlier observations in the Groningen ART cohort study at younger ages135,140,172 and several other studies on long-term follow-up on the neurodevelopmental outcome after ART.104-108,194,195

Our study also indicates that the underlying causes of subfertility do not seem to affect neurological development of 4-year-olds, again corroborating our findings at the age of 2 years.172 We were also unable to find specific associations between the presence of subfertility per se and the adverse neurological outcome. Most likely, the absence of a difference in the outcome between children born to subfertile parents and those born to fertile parents can be explained by the composition of the reference group. The only time when the neurological outcome of our subfertile ART cohort was significantly worse than that of the fertile reference group was in the study at 3 months. In this study the reference group consisted of a representative sample of the general population, as the infants had been assessed as a part of a general health check-up provided for all infants.135 At the follow-up at 2 years, and currently at 4 years – when we did not find a significant neurological disadvantage for the subfertile group – the fertile reference group consisted of children whose parents volunteered for participation in the study. This may have caused selection bias, as parents who are more concerned about the health and development of their children are presumably more willing to participate in a neurodevelopmental outcome study.196 The idea of selection bias is supported by the relatively high proportion of children in the fertile reference group with complex MND (14%), a proportion that is higher than that reported in the general population (6-7%).183

5


83

Interestingly, we found that a longer TTP was associated with a higher prevalence of complex MND, which was particularly true for girls. The finding that the severity of subfertility may be the decisive factor rather than the presence of a history of subfertility per se corroborates previous reports (mild psychomotor delay at 18 months;97 modest increased risk for DCD in 7-year-olds;98 neurological outcome at 2 years191). Others suggested the presence of a sex-specific effect of IVF and ICSI on the neurological outcome,100,108,192 with boys having a less favourable neurodevelopmental outcome than girls. In general, boys have twice as often complex MND as girls,90,183 which may be related to the sex-specific timing of developmental events in the nervous system.197-200 In our subfertile study group a relatively high proportion of children, boys as well as girls, had complex MND (total group: 24%; boys: 21% girls: 28%), a proportion that is considerably higher than in the general population (6-7%).183 Our data may imply, that a history of prolonged subfertility, abolishes the sex-specific neurodevelopmental vulnerability of the boys, by boosting vulnerability for both sexes. The negative effect of TTP, particularly in girls, may be due to the stress which subfertile couples may experience when finally a pregnancy is achieved. As a result the fetus of subfertile couples may be exposed to prenatal stress. This prenatal stress may exert a long-term physiological and neurodevelopmental effect, notwithstanding the fact that couples with a relatively long TTP generally have little trait anxiety and a good mental health, when the child has reached the age of 1 year.201 Yet, the latter does not preclude that the offspring is affected by stress in utero. Moreover, subfertile couples may underestimate the incurred stress as the joyfulness of having a child at last may overshadow the rough period of prolonged subfertility.202,203 Evidence is accumulating that maternal stress during pregnancy is associated with an increased risk of disturbance in offspring neurodevelopment.204,205 Cortisol and testosterone have been proposed as mediating hormones between maternal mental status and fetal development, as cortisol and testosterone are both the end products of two hormonal axes that can interact with each other, and perinatal exposure to high levels of cortisol and testosterone is associated with perinatal adversities and health problems. It is well-known that gonadal steroids play a crucial role in the sexual differentiation of the brain and its development. Animal studies have shown that prenatal stress masculinizes females and feminizes males, under the influence of androgens, impairing reproductive capacity and reducing sex-specific behaviour.206-210 Similar associations have been reported in humans by the studies of Barrett et al.: exposure to prenatal stress was associated with a longer, more masculinized anogenital distance (AGD – an indicator of prenatal androgen exposure) and masculinized play behaviour in girls; the feminizing effect of prenatal stress in males was less clear.211,212 The Barrett-findings support our hypothesis that a longer TTP may be associated with exposure to in utero stress, which in turn may masculinize the female brain and thus induce a more boy-like neurological vulnerability in girls.

The present study did not demonstrate associations between fertility-related factors and neurological optimality in terms of movement fluency and NOS. The question arises why a longer TTP was associated with clinical neurological condition in terms of complex

5

84

MND and not with the derived parameters fluency score and NOS. Clinical neurological condition and neurological optimality are highly correlated; however, a reduction in neurological optimality does not necessarily mean the presence of a neurological deviation, as the range for optimality is narrower than that for normality.160 Moreover, the classification of MND involves a threshold: MND is only present if a certain number of deviant neurological signs are present. This means that the severity of subfertility is not associated with milder degrees of neurological non-optimality, but only has an impact on the risk for the clinically relevant form of MND.

Strengths and limitations The major strength of the present study is its design, which offered the opportunity to study the effects of several components of assisted conception such as ovarian hyperstimulation and the in vitro procedure, on the neurological outcome of offspring of subfertile parents. The prospective design of our study reduced potential selection bias based on the child’s development or health, since we invited the couples of the Groningen ART cohort in the third trimester of pregnancy. The minimal postnatal attrition rate of 9.3% and the blinding of our assessors to the mode of conception of the Groningen ART cohort add to the strength of our study.

Another strength of the study is the use of age-specific and highly sensitive measurements to study the neurological outcome in 4-year-olds. Subtle changes in neurological development are relatively unimportant on an individual level; however, on a population level they are relevant, since ART children represent a substantial part of society. Most studies however use relatively gross measurements on the neurological outcome, such as the Bayley’s or Griffiths Scale of Infant Development. For example, Bouwstra et al. demonstrated an adverse effect of neonatal trans-fatty acid status on neurological development with the use of the Hempel assessment but not with the use of the Bayley’s Scale of Infant Development.164

A number of caveats need to be discussed regarding the present study. First, the results cannot be generalized to multiples, as we studied singletons only. Being a member of a twin is associated with an increased risk for developmental problems,179 and ART is known to be associated with multiple births. Secondly, the original power calculation was based on the neurological outcome in the ART study groups and not based on possible associations between TTP and the neurological outcome. We performed a post hoc power analysis for the outcome parameter complex MND related to TTP. For the total group of children, the analysis indicated that the current effect size of 1.207 related to the current sample size of 195 children equals a power of 67.7%. This means that the sample size of the total group is somewhat underpowered to detect a relatively small OR of 1.207. Although our finding is plausible from a neurobiological point of view, the result has to be interpreted with caution, given the chance of a false-positive finding. Due to the larger effect size calculated for the girls, the size of the subgroup of girls (n=94) was large enough to detect relevant effects: the effect size of 1.542 was detectable with a power of 84.0%. Another limitation of the

5


85

study is the composition of the fertile reference group which does not match the way in which we recruited the children of the Groningen ART cohort. The reference group was newly recruited for the 4 year follow-up period. As a consequence it was not possible to blind the assessors to the fertility status of the reference group. Moreover, the composition of the reference group may have caused selection bias: parents who were more concerned about the health and development of their children are presumably more willing to participate in a neurodevelopmental outcome study.196

In conclusion, our study suggests that ovarian hyperstimulation, the in vitro procedures, a combination of these two factors, the presence of subfertility and specific causes of subfertility are not associated with the worse neurological outcome at 4 years of age. Yet, the study also suggests that increased TTP is associated with a higher prevalence of complex MND, particularly in girls. This suggests that rather the severity of subfertility than its presence affects the neurological outcome, an effect which is especially expressed in girls: severe subfertility induces in girls a male vulnerability to develop MND. Our results underline the notion of the importance of long-term follow-up of development and growth of children born after ART, especially since in society maternal age at child birth, subfertility and the application of ART are steadily increasing.

ACKNOWLEDGEMENTS We thank participating parents and children for their cooperation and enthusiasm during the assessments; Arend F. Bos, M.D. Ph.D., for his help in initiating the study and Maaike Haadsma, M.D. Ph.D., for her help in including the participants; Rosan Aapkes, M.D., Elise Bennink, M.Sc., Hanneke Geut, M.D., Marjolein Jongbloed-Pereboom, M.Sc. and Bertine de Vries, M.Sc., for support in collecting and organizing the data; Michiel Schrier, M.Sc., Linze Dijkstra B.Sc., and Loes de Weerd for technical assistance.

5

86

Subfertility factors rather than assisted conception factors affect cognitive and behavioural development of 4-year-old singletons

87

CHAPTER 6 Subfertility factors rather than assisted conception factors affect cognitive and behavioural development of 4-year-old singletons

Pamela Schendelaar


Maas Jan Heineman

Karin J. Middelburg

Jorien Seggers


Mijna Hadders-Algra

Submitted

88

ABSTRACT Objective To study underlying causal relationships between ovarian hyperstimulation, in vitro procedures, subfertility components and child cognition and behaviour. Design Longitudinal, prospective follow-up study. Setting Academic Centre. Patients Four-year-old singletons born to subfertile couples (subfertile group, n=195), including singletons born after ovarian hyperstimulation IVF (COH-IVF, n=63), modified natural cycle IVF (MNC-IVF, n=53) and natural conception (Sub-NC, n=79). Newly recruited 4-year-old singletons born to fertile parents after natural conception (reference group, n=98). Intervention None. Main outcome measures Primary outcome: total intelligence quotient (IQ). Behavioural outcome: total problem T-score. Causal inference search algorithms and structural equation modelling were applied as statistical tools. These are unlike traditional statistics able to unravel underlying causal mechanisms and distinguish between confounders and intermediate effects. Results No direct or indirect underlying causal effect was found between ovarian hyperstimulation and the in vitro procedure and child cognition and behaviour. However, direct negative causal effects were found between a) severity of subfertility (time to pregnancy, TTP) and cognition and b) presence of subfertility and behaviour. These associations were confounded by maternal age at child conception and maternal educational level (GES search algorithm, penalty discount: 0.5; model fit: X2 = 84.7, df = 72, P = 0.145, X2/df = 1.176; BIC-score = -317.0). Conclusion Our study demonstrated no direct or indirect causal effects between ovarian hyperstimulation or in vitro procedures and cognitive and behavioural outcome in 4-year-old singletons born to subfertile couples. However, suffering from subfertility, especially from more severe subfertility – which by itself is associated with higher maternal age and educational level – negatively affects child cognition and behaviour.

6


89

INTRODUCTION Long-term follow-up of the increasing number of children born following assisted reproductive techniques (ART) is important. ART is associated with perinatal adversities,52-56 yet, ART does not seem to affect cognitive and behavioural development during the first postnatal years.89 However, children may grow into developmental deficits since it takes time for developmental disorders to emerge.90

Various ART-related factors could potentially interfere with development, such as ovarian hyperstimulation,171 the in vitro procedures,31 the underlying subfertility59-62 and parental characteristics.7,57,58

Results of long-term studies on cognitive and behavioural development in ART children vary, partly due to difficulties in distinguishing relationships between ART and underlying characteristics of subfertility, parents and child. Cognitive and behavioural outcome of ART children has been reported as similar,82,108,109,111,213 to worse86,97,104,113-115,126,214 or better107,112 than that of naturally conceived children.

In order to investigate the influence of specific factors involving assisted conception on cognitive and behavioural outcome, we composed the Groningen ART cohort. The cohort consists of three groups of singletons born to subfertile couples following conventional controlled ovarian hyperstimulation in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) (COH-IVF), modified natural cycle IVF/ICSI (MNC-IVF) and natural conception (Sub-NC).135 We previously reported no differences in neurological, cognitive and behavioural outcome between the groups up to age 4.140,141,172,215 However, at 2 and 4 years, time to pregnancy (TTP) was associated with a less optimal neurological condition.191,216 Additionally, at 2 years the cohort children had higher scores on anxious-depressed behaviour than children in a reference group with fertile parents.140,141,172,215 This suggests that rather subfertility than ART components affect neurodevelopment. At 4 years we applied a causal inference approach to evaluate factors affecting cardiometabolic outcome. It indicated direct positive effects of COH-IVF and not of MNC-IVF on systolic blood pressure percentiles and subscapular skinfold thickness, suggesting that ovarian hyperstimulation was involved in worse cardiometabolic outcome in our 4-year-old IVF singletons.217

In analogy, the primary aim of the present study is to explore the underlying causal relationships between ovarian hyperstimulation, the in vitro procedure and the combination of both and cognition and behaviour in 4-year-old singletons. Secondly, we aim to explore the underlying causal relationships between several aspects of subfertility and cognition and behaviour. We addressed the contribution of a) the presence of a history of subfertility by including a reference group of 4-year-old singletons born to fertile parents, and b) the duration of subfertility in terms of TTP, as a proxy for the severity of subfertility. We have chosen cognitive outcome as our primary outcome parameter as cognition has a stronger neurobiological basis than behaviour.190,218

6

90

MATERIAL AND METHODS

Participants Couples who achieved a singleton pregnancy following IVF/ICSI with a term date between March 2005 and December 2006 were recruited at the Department of Reproductive Medicine of the University Medical Center Groningen (UMCG) and were invited for participation during their third trimester.141 This resulted in two groups of children born following controlled ovarian hyperstimulation IVF/ICSI (COH-IVF) and born following modified natural cycle IVF/ICSI (MNC-IVF).136,137 Couples treated with donated oocytes or cryopreserved embryos were excluded. A third group was formed by naturally conceived children born to subfertile couples (Sub-NC) who had tried to conceive for at least one year.

A new retrospective reference group was recruited between December 2009 and February 2012 at six local child welfare centres. Parents of 4-year-old singletons who visited these centres for routine general health care were invited to participate. Couples who had tried to achieve pregnancy for more than one year or achieved pregnancy by any form of assisted conception were excluded.

Setting Prenatal, perinatal and demographic information was collected two weeks postterm.135 Information on the causes and treatment of infertility was retrieved from medical records.

The follow-up assessments were carried out by trained assessors supervised by a neurodevelopmental expert (M.H.-A.), who were blind to the mode of conception. Blinding was not possible for the reference group, as this group was recruited separately from the subfertile groups. The assessments were carried out between February 2009 and February 2012 at the UMCG.

Measurements Cognitive development Cognitive development was evaluated using the Kaufman Assessment Battery for Children, second edition (K-ABC-II).166 This standardized instrument measures cognitive and processing abilities in children aged 3 to 18 years. Outcome is expressed in a total intelligence quotient (IQ) score and four IQ scale scores. In the present paper only the total IQ score is used. Raw test scores are normalized into global scores (mean: 100, standard deviation [SD]: 15). Reliability and validity of the K-ABC-II are good.166 The original American norms were applied as Dutch norms are lacking.

Behavioural development Behavioural development was evaluated using the validated Dutch version of the Child Behavior Checklist (CBCL).167,168 The CBCL is a parental questionnaire to identify emotional and behavioural problems in 1.5 to 5-year-olds, classified into problem scales, such as anxious/depressed and attention problems. The sum of all questions results in the total

6


91

problem scale score. Raw test scores are normalized into T-scores (mean: 50, SD: 10). Higher T-scores represent more problematic behaviour. The reliability and validity of the CBCL are good.167

Statistical analysis Fisher’s exact tests, Mann-Whitney U-tests or Student’s t-tests were performed to investigate differences between groups using the IBM Statistical Package for the Social Sciences (SPSS), version 20. P-values < 5% were considered statistically significant.

Two separate sets of explorative analyses – each consisting of a dozen of explorations – using causal inference search algorithms were performed: one focusing on ART treatment effects (ovarian hyperstimulation; the in vitro procedure; combination of both), the other focusing on subfertility effects (presence and severity of subfertility, recorded as TTP in half-years) on developmental outcome, while taking into account causal relations of these factors and other parental and child aspects.

Causal inference search algorithms result in (classes of) causal models that are found to be compatible with the data, based on the theory of causal graphs.219,220 A causal model consists of a model and a graph, in which the latter describes the causal relation between the variables and consists of vertices (the variables), connected by edges, which can be oriented by arrows.219 An oriented edge represents a direct causal effect between the two connecting vertices.221

We applied Conservative Peter-Clark (CPC), Greedy Equivalency Search (GES) and Conservative Fast Causal Inference (CFCI) algorithms.217 The algorithms differ in search approach and data assumptions and use particular thresholds (GES algorithm, penalty discounts: 0.4 to 2; CPC and CFCI algorithm, alpha values: 0.05 to 0.4) of announcing a certain effect significant.222 All search algorithms were applied while taking into account background knowledge which was logically dictated by time constraints. The resulting graphs per se represent an observationally equivalent class of causal models, i.e. other graphs from that class are equally likely to have generated the data.217 Model fit was calculated and compared between the found classes of models, by choosing one directed acyclic graph, representing such a class. We used the X2 test, X2/degrees of freedom (df) and the Bayesian information criterion (BIC) to compare models.223 Additionally, structural equation modelling was applied to a graph that represents one of the better fitted classes of models to estimate effect sizes.223,224

By comparing the various graphs resulting from the performed search algorithms, the most prominent and consistent direct effects could be distinguished. Additionally, effects that are consistently not found can be considered an indication of the absence of a causal effect.220 The explorative analyses were performed using the freeware program TETRAD, version 403.10-6225 and the sem library in R, version 2.15.0.226

6

92

Ethical approval The Medical Ethical Commission of the UMCG approved the study design. Parents provided written informed consent for study participation of their child.

FIGURE I. Flow chart of the Groningen ART cohort.

6


93

RESULTS

Ovarian hyperstimulation and the in vitro procedure: the three subfertile groups Participation and demographic characteristics During the prenatal period, 89 COH-IVF children, 79 MNC-IVF children and 143 Sub-NC children were eligible for participation of which respectively 68 (76%), 57 (72%) and 90 (93%) children were included in the study.135 Generally, social, obstetrical and neonatal characteristics of participants and non-participants were similar, except for a lower maternal age in non-participating Sub-NC mothers compared to participating Sub-NC mothers (P = 0.030).135

At 4 years, 5 (7%) COH-IVF children, 4 (7%) MNC-IVF children and 11 (12%) Sub-NC children were lost to follow-up (total postnatal attrition rate 9.3%, Figure I). One MNC-IVF girl died at 3 weeks after birth because of a congenital heart disorder. Intra-group characteristics for participants and non-participants were similar, except for a longer TTP in participating MNC-parents compared to non-participating MNC-parents (P = 0.021). The demographic characteristics of the three groups are listed in Table I.

Cognitive and behavioural development Two Sub-NC children had missing data on the K-ABC-II. Eventually, data of 193 children (63 COH-IVF, 53 MNC-IVF and 77 Sub-NC) were analysed. The majority of children had total IQ scores within the normal range, except for two COH-IVF children (IQ: 82 and 79) and two MNC-IVF children (IQ: 79 and 77). The mean total IQ scores of the three ART study groups are listed in Table II.

One COH-IVF, one MNC-IVF and one Sub-NC child had missing data on the CBCL. Eventually, data of 192 children (62 COH-IVF, 52 MNC-IVF and 78 Sub-NC) were analysed. The majority of children of the ART study groups had total problem T-scores within the normal range, except for two COH-IVF children and 4 Sub-NC children. The mean total problem T-scores of the three ART study groups are listed in Table II. Causal inference approach The causal graph that had the best model fit was the causal model found as a result of running the GES algorithm with a penalty discount of 0.5 (X2= 98.5, df=94, P = 0.356, X2/df= 1.048, BIC-score= -389.1). This model indicated the absence of direct effects of COH-IVF or IVF on cognitive or behavioural outcome given the other variables in the model. A direct effect of COH-IVF or IVF was also absent in causal graphs resulting from other searches (data not provided).

6

94

COH - IVF MNC - IVF Sub - NC Subfertile group Reference groupn = 63 n = 53 n = 79 n = 195 n = 98

Child characteristicsMale gender, n (%) 34 (54.0) 26 (49.1) 41 (51.9) 101 (51.8) 54 (55.1)First born, n (%) 43 (68.3) 37 (69.8) 49 (62.0) 129 (66.2)** 46 (46.9)**

Corrected age at examination at 4 years of age (months), median (range) 50 (47.5 - 60.1) 48.9 (48.0 - 52.5) 48.9 (47.9 - 56.4) 48.9 (47.5 - 60.1) 49.1 (48.0 - 54.6)

Birth characteristicsGestational age (weeks), median (range) 39.4 (33.4 - 42.3)* 40.1 (34.6 - 42.6) 40.0 (30.1 - 42.7)* 40.0 (30.1 - 42.7) 40.1 (32.0 - 42.4)Preterm birth (< 37 weeks), n (%) 7 (11.1) 6 (11.3) 5 (6.3) 18 (9.2) 3 (3.1)Birthweight (grams), mean (SD) 3393.1 (563.2)* 3384.4 (585.7)* 3577.9 (519.4)*/* 3465.6 (557.2)* 3599.7 (507.0)*

Low birthweight (< 2500 gram), n (%) 3 (4.8) 4 (7.5) 3 (3.8) 10 (5.1) 2 (2.1)Small for gestational agea, n (%) 0 (0) 3 (5.7) 1 (1.3) 4 (2.1) 4 (4.3)Signs of fetal distressb, n (%) 19 (30.2) 15 (28.3) 34 (43.0) 68 (34.9) 29 (29.6)

Neonatal characteristicsApgar score 5 min < 7e, n (%) 0 0 1 (1.3) 1 (0.5) 3 (4.2)Neonatal intensive care admission, n (%) 1 (1.6) 2 (3.8) 5 (6.3) 8 (4.1) 10 (10.2)

Parental characteristicsMaternal age at conception in years, median (range) 32.3 (26.3 - 40.9) 32.8 (25.3 - 37.5) 33.0 (22.2 - 40.3) 32.8 (22.2 - 40.9)*** 30.4 (18.8 - 40.5)***

Paternal age at conception in yearse, median (range) 35.4 (27.5 - 56.1)* 34.0 (28.3 - 47.8)* 35.0 (25.5 - 48.7) 35 (25.5 - 56.1)*** 32.6 (22.5 - 45.1)***

Education level mother (highc), n (%) 20 (31.7) 20 (37.7) 37 (46.8) 77 (39.5)* 52 (53.1)*

Education level father (highc)e, n (%) 28 (46.7) 17 (32.7) 29 (36.7) 74 (38.7)** 53 (56.4)**

Smoking during pregnancy, n (%) 7 (11.1) 7 (13.2) 9 (11.4) 23 (11.8) 5 (5.1)Alcohol consumption during pregnancy, n (%) 3 (4.8) 0 2 (2.5) 5 (2.6) 4 (4.1)

Divorced, n (%) 1 (1.6) 1 (1.9) 3 (3.8) 5 (2.6)** 11 (11.3)**

Fertility parametersTime to pregnancy in yearsd,e, median (range) 4.0 (0.0 - 13.5)*** 4.0 (0.5 - 13.5)*** 2.5 (0.5 - 11.5)***/*** 3.5 (0.0 - 13.5)*** 0.5 (0.5 - 1.0)***

Type of infertility (primary), n (%) 35 (55.6) 32 (60.4) 41 (51.9) 108 (55.4) n.a.ICSI performed, n (%) 41 (65.1) 26 (49.1) n.a. n.a. n.a.

c University education or vocational colleges.

Characteristics

d TTP was recorded in half-years. In case of a miscarriage the onset of TTP restarted, therefore TTP may be shorter than one year.e Overall missing data: gestational age n=1, preterm n=1, birthweight n=4, low birthweight n=3, small for gestational age n=1, Apgar score 5 min < 7 n=31, maternal age at conception n=1, paternal age at conception n=8, education level father n=8, time to pregnancy n=1.

TABLE I. Characteristics of participating parents and children.

Mann-Whitney U -tests, Student's t -test or Fisher's exact tests were performed to investigate differences between groups. * P < 0.05; ** P < 0.01; *** P < 0.001. Note: COH-IVF: infants born following controlled ovarian hyperstimulation IVF or ICSI, MNC-IVF: infants born following modified natural cycle IVF or ICSI, and Sub-NC: naturally conceived infants born to subfertile parents. In the subfertile group COH-IVF, MNC-IVF and Sub-NC groups are taken together. a Birth weight for gestational age is < -2 standard deviations compared with the Dutch reference population (Dutch reference tables, perinatal Registration Netherlands).b Signs of fetal distress denoted by meconium stained amniotic fluid and/or cardiotocographic signs and/or acidosis.

6


95

Sub-

NCMN

C-IV

FCO

H-IV

FRe

feren

ce g

roup

Subf

ertile

gro

upn

= 77

n = 5

3n

= 63

n = 9

7n

= 193

Cogn

itive

dev

elop

men

t

Refe

renc

e, m

ean

(SE)

Refe

renc

e, m

ean

(SE)

Mean

diff

eren

ce [

95%

CI]

Total

IQ10

8.9 (1

.22)

-3.7

[-7.8

; 0.5]

-2.8

[-6.8

; 1.2]

110.4

(1.10

)-3

.5 [-6

.3 ; -

0.7]

Sequ

entia

l IQ95

.9 (1

.31)

1.2 [-

3.0 ;

5.4]

0.5 [-

3.5 ;

4.5]

99.6

(1.18

)-3

.2 [-6

.0 ; -

0.3]

Simult

aneo

us IQ

115.5

(1.59

)-1

.5 [-6

.3 ; 3

.4]-2

.4 [-7

.0 ; 2

.2]11

7.8 (1

.46)

-3.5

[-6.9

; -0.0

] Le

arnin

g IQ

98.3

(1.26

)-2

.6 [-6

.7; 1.

5]-1

.5 [-5

.4 ; 2

.4]99

.6 (1

.02)

-2.4

[-5.2

; 0.3]

Know

ledge

IQ11

2.7 (1

.12)

-6.0

[-10.3

; -1

.8]-4

.4 [-8

.5 ; -

0.4]

111.3

(1.26

)-1

.7 [-4

.7 ; 1

.4]

Sub-

NCMN

C-IV

FCO

H-IV

FFe

rtile

refe

renc

e gr

oup

Subf

ertil

e gr

oup

n =

78n

= 52

n =

62n

= 96

n =

192

Beha

viou

ral d

evel

opm

ent

Refe

renc

e, m

ean

(SE)

Refe

renc

e, m

ean

(SE)

Mean

diff

eren

ce [

95%

CI]

CBCL

total

prob

lem T

-scor

e47

.8 (1

.06)

-1.6

[-4.8

; 1.5]

-0.5

[-3.4

; 2.4]

44.4

(0.99

)2.2

[-0.0

; 4.5

]CB

CL in

terna

lizing

T-sc

ore

48.0

(1.14

)-1

.3 [-4

.8 ; 2

.3]-3

.0 [-6

.3 ; 0

.4]45

.0 (1

.04)

1.8 [-

0.7 ;

4.2]

CBCL

exter

naliz

ing T

-scor

e48

.7 (1

.01)

-1.2

[-4.2

; 1.9]

-0.5

[-3.4

; 2.4]

46.2

(0.95

)2.0

[-0.2

; 4.2

]No

te: C

OH-IV

F: in

fants

born

follo

wing

contr

olled

ovar

ian hy

perst

imula

tion I

VF or

ICSI

, MNC

-IVF:

infan

ts bo

rn fo

llowi

ng m

odifie

d natu

ral c

ycle

IVF

or IC

SI, a

nd S

ub-N

C: na

turall

y con

ceive

d infa

nts bo

rn to

subfe

rtile p

aren

ts. In

the s

ubfer

tile gr

oup C

OH-IV

F, M

NC-IV

F an

d Sub

-NC

grou

ps ar

e tak

en to

gethe

r. CB

CL: C

hild B

ehav

ior C

heck

list.

TABL

E II.

Grou

p di

ffere

nces

in IQ

and

CBCL

scor

es o

f par

ticip

atin

g ch

ildre

n.

Mean

diff

eren

ce [9

5% C

I]

Mean

diff

eren

ce [9

5% C

I]

6

96

Presence and severity of subfertility: inclusion of the fertile reference group Participation and demographic characteristics Parents of 215 eligible singleton reference children were invited to participate in the present follow-up study. Ninety-eight (46%) parents allowed their child to participate in the reference group. Demographic characteristics such as sex, gestational age, firstborn, maternal age and parental educational levels were similar for participants and non-participants of the reference group (data not provided). Demographic characteristics of participants of the subfertile and reference groups are listed in Table I. Cognitive and behavioural development Two children of the subfertile group and one reference child had missing data on the K-ABC-II. Eventually, data of 290 children (subfertile group: 193, reference group: 97) were included in the analyses. Four children of the subfertile group had impaired total IQ scores, whereas none of the reference children had. The mean total IQ scores of the groups are listed in Table II.

Three children of the subfertile group and two reference children had missing data on the CBCL questionnaire. Eventually, data of 288 children were included in the analyses. The majority of children in both groups had total problem T-scores within the normal range, except for six children of the subfertile group and four reference children. The mean total problem T-scores of the subfertile group and the reference group are listed in Table II.

Causal inference approach The causal graph that presented the class of found models with the best model fit as a result of the causal effect search algorithms is represented in Figure II (X2= 84.7, df=72, P = 0.145, X2/df=1.176, BIC-score=-317.0). It indicates that the presence of subfertility did not have a direct or indirect effect on cognition. Similar results were found in resulting causal graphs from other searches (data not provided). The causal graph suggests that cognitive outcome was directly affected by two other factors: TTP and maternal smoking during pregnancy. A longer TTP was associated directly with a lower total IQ score. This effect was confounded by high maternal education through various causal paths: one path ran from maternal education via maternal and paternal age, via TTP to IQ, whereas another path ran from maternal education via smoking during pregnancy to IQ. The other direct effect on the child’s total IQ score was the negative effect of maternal smoking during pregnancy, an effect that again was confounded by maternal education. Taking this complexity of the causal pathways into account and correcting for maternal education level, the direct, unconfounded effect of TTP on total IQ score is estimated to be -0.712 (Figure II).

The causal graph suggests that the presence of subfertility had a direct adverse effect on behavioural outcome (Figure II). As such, subfertility status acted as mediator on three different paths running from high maternal educational level on total problem T-score. These paths ran either via maternal age at child conception, or via TTP, or via a combination of maternal age and TTP, including an effect of paternal age at child conception (Figure II).

6


97

It also indicates that maternal age directly affected behavioural outcome: a higher maternal age was associated with less behavioural problems in the child. Also in this path, maternal age mediated the effect of high maternal educational level. Subsequently, maternal age at conception and high maternal educational level are considered to be confounders for the effect of subfertility status on the total problem T-score. The unconfounded, direct effect of the presence of subfertility on the total problem T-score is estimated as follows: conditionally on maternal age at conception, high maternal educational level and/or TTP, the direct effect of subfertility status is estimated to be 3.003 for less optimal behaviour.

The causal graph also suggests a relationship between the two outcome parameters, cognition and behaviour: an increase in total IQ score with one point was associated with a decrease in total problem T-score with 0.120 points, indicating better behaviour in case of a more optimal cognitive development (Figure II). According to the causal implications, TTP has an indirect (mediated by subfertility status and total IQ score) effect on more problematic behaviour (Figure II).

DISCUSSION The present study indicates that in 4-year-old singletons born to subfertile couples direct or indirect causal effects between ovarian hyperstimulation or the in vitro procedure on the one hand and cognitive and behavioural outcome on the other hand are absent. However, a direct causal relationship was found between the severity of subfertility (TTP) and cognition at age 4. Also, a direct negative causal relationship between the presence of subfertility and behaviour was found. Both effects were confounded by maternal age at child conception and maternal educational level. TTP had a direct negative effect on the child’s cognitive outcome, whereas the presence of subfertility had a direct adverse effect on behaviour. Given the finding that cognition and behaviour are directly related to one another (with behaviour being influenced by cognition) our results suggest that the severity of subfertility has an indirect – through subfertility status as well as through IQ – effect on the child’s behavioural scores, i.e. was associated with more behavioural problems.

Our findings strengthen the notion that aspects of subfertility rather than components of ART are involved in determining cognitive and behavioural outcome of children born after IVF/ICSI. Our statistical approach also revealed important confounders for these effects, especially maternal age and education. In turn, the effect of the presence of subfertility on behavioural outcome was confounded by TTP. Our results imply that suffering from subfertility per se – especially from more severe subfertility, which in turn is affected by a higher maternal age and higher educational level – negatively affects the child’s cognitive and behavioural outcome.

6

98

The fact that already known and generally expected mechanisms were also detected (Figure II), underlines the general robustness and validity of our results. Not only the causal graph with the best model fit of the GES search algorithm, but also other graphs resulting from the GES searches with a good model fit showed similar mechanisms concerning the abovementioned effects, indicating a certain consistency of the found relations between the variables. Similar effects regarding maternal age, maternal educational level and TTP on the outcomes were also found with the CPC search algorithm (model fit indices for the CPC-result with the best model fit: alpha value 0.15; model fit: X2= 72.1, df=68, P=0.344, X2/df= 1.060 and BIC-score=-307.3) The fact that similar mechanisms are revealed by two different algorithms in terms of search approach and data assumptions, underline the validity of found mechanisms. One might have expected to find direct effects of birthweight and gestational age on cognitive and behavioural outcome, however, our analyses did not result in such effects (Figure II). Birthweight is indeed associated with the child total problem T-score via maternal age at conception, but this association disappeared after correction for

6

FIGURE II. Causal graph depicting the underlying causal mechanism as found in the subfertile (COH-IVF, MNC-IVF and Sub-NC groups pooled) and fertile reference group regarding presence and severity of subfertility and cognitive and behavioral development The causal graph representing the model with the best model fit of the causal inference search algorithms (GES search algorithm, penalty discount: 0.5; model fit: X2= 84.7, df=72, p=0.145, X2/df=1.176 and BIC-score=-317.0). Each arrow in the graphs is accompanied by two numbers, reflecting the size of the causal effect per unit change. The first number is the estimated regression weight and the second number between brackets is its corresponding standard error. A positive estimated regression weight in the graph means that higher values of the variable are associated with an increase in the value of the variable to which the arrow is pointing. Likewise, a negative estimated regression weight in the graph means that higher values of the variable are associated with a decrease in the value of the variable to which the arrow is pointing. TTP was recorded in half-years. High educational level indicates university education or vocational colleges.


99

maternal age. This is in line with our expectations: our groups contain few preterm and low birthweight infants. Therefore, it seems more probable that our analyses resulted in effects of subfertility and maternal factors on developmental outcome, rather than in effects of birthweight and gestational age.

To our knowledge, this is the first study in the context of neurodevelopment in ART children using causal inference search algorithms combined with structural equation modelling. With that, this is the first study that is able to distinguish between direct and indirect causal effects and to detect and correctly adjust for confounders at the same time. Our findings contribute to the clarification of the current evidence on long-term cognitive and behavioural outcome of children born following ART;86,97,104,113-115,126,214 an evidence that still needs further proof. Our findings that the severity of subfertility plays a role in cognitive development is in line with Zhu et al., who reported that longer TTP may be associated with a delay in achieving certain milestones, in particular those involved in cognitive and language development.97 However, in another study by Zhu et al. the authors were not able to demonstrate a subfertility effect on child behaviour.213

To intelligibly answer our research questions, we performed two separate sets of explorative analyses: one to zoom in on effects of ART treatment, the other to focus on the effects of presence and severity of subfertility on developmental outcome. The first exploration can only be translated to its particular (smaller) subfertile population and its findings do not extend to the different population that was used for the second exploration. In the latter analysis a reference group was added, allowing for the in concert evaluation of the effect of the presence of subfertility and that of its severity.

The explorative analyses were restricted to a certain amount of variables which we have chosen a priori based on literature. Moreover, the search algorithms are particularly suitable for continuous values and less for categorical values. We have performed several analyses with the variables parity, the presence of siblings, nursery and bilingualism taken into account. These variables did not contribute to the model and would not have altered the interpretation of the models rather than unnecessarily complicate it. The same held true for the variable ICSI. We primarily aimed to explore the underlying causal relationships between ovarian hyperstimulation and the in vitro procedure and developmental outcome and we were not primarily interested in the effect of ICSI in addition to IVF.

Besides the additional value of the statistics applied, the present study has some other important strengths. Our study design enabled us to study the effects of separate components of assisted conception such as ovarian hyperstimulation, the in vitro procedure and subfertility-related aspects on the child’s cognitive and behavioural outcome. Additional strengths of the ART cohort part of the study are the attrition rate of 9.3%, the blinding of assessors to the mode of conception and the prospective design.

It must be realized that the statistical tools used in our study are explorative in nature and especially serve as indications for new research hypotheses, meaning that our results need to be interpreted with appropriate caution.217,227 However, this caution is not restricted to search algorithms alone: applying sets of multivariable regression analyses is in fact also

6

100

explorative in nature. Due to the explorative nature of our analyses a post hoc power calculation is irrelevant. Previously group sizes of the Groningen ART cohort was based on neurological outcome at 18 months.140 As a general rule for a structural equation model, ten or more observations for each variable may be considered reliable, a criterion that we have met.

A more general limitation of the study is the composition of the reference group which we recruited retrospectively. Consequently, we were not blinded to group status of the reference children. Parents who are concerned about their child’s health in general are more likely to participate in developmental studies. However, this means that the associations reported in the present study may have been underestimated. The relatively low participation rate in the reference group may be an indicator for selection bias.

Another issue is that our results cannot be generalized to multiples. Being a member of a twin is associated with an increased risk for developmental problems,179 and ART is known to be associated with multiple births.

To conclude, our study implies that suffering from subfertility per se, and especially from more severe subfertility – which by itself is affected by higher age and high educational level of the mother – negatively affects the child’s cognitive and behavioural outcome. Long-term follow-up of development and growth of children born from subfertile couples, in particular when ART is applied, remains important, given the steadily increasing prevalence of subfertility in modern society.

ACKNOWLEDGEMENTS

We thank participating parents and children for their cooperation and enthusiasm during the assessments; Arend F. Bos, M.D. Ph.D., for his help in initiating the study and Maaike Haadsma, M.D. Ph.D., for her help in including the participants; Rosan Aapkes, M.D., Elise Bennink, M.Sc., Hanneke Geut, M.D., Marjolein Jongbloed-Pereboom, M.Sc. and Bertine de Vries, M.Sc., for support in collecting and organizing the data; Michiel Schrier, M.Sc., Linze Dijkstra B.Sc., and Loes de Weerd for technical assistance.

6

The effect of preimplantation genetic screening on neurological, cognitive and behavioural development in 4-year-old children: follow-up of a RCT

101

CHAPTER 7 The effect of preimplantation genetic screening on neurological, cognitive and behavioural development in 4-year-old children: follow-up of a RCT

Pamela Schendelaar

Karin J. Middelburg

Arend F. Bos

Maas Jan Heineman

Joke H. Kok


Jorien Seggers


Mijna Hadders-Algra


102

ABSTRACT

Study question Does embryo biopsy inherent to preimplantation genetic screening (PGS) affect neurological, cognitive and behavioural development of 4-year-old children? Summary answer PGS does not seem to affect neurological, cognitive and behavioural development of 4-year-old singletons, however, our data suggest that it may be associated with altered neurodevelopment in twins. What is known already Evidence concerning the safety of PGS on neurodevelopmental outcome in offspring is scarce. The present study provides information on neurodevelopmental, cognitive and behavioural outcome of 4-year-old PGS offspring. Study design, size, duration A prospective, assessor-blinded follow-up study of children born to women who participated in a multicentre RCT on the effect of IVF with or without PGS. Participants/materials, setting, methods At 4 years, 49 children (31 singletons, 9 sets of twins) born following IVF with PGS and 64 children (42 singletons, 11 sets of twins) born following IVF without PGS (controls) were assessed (postnatal attrition 18%). Neurological development was evaluated with the standardized, age-specific and sensitive neurological examination according to Hempel, resulting in a neurological optimality score (NOS), a fluency score and the rate of adverse neurological outcome. Primary outcome was the fluency score, as fluency of movements is easily reduced by subtle dysfunction of the brain. Cognitive development was evaluated with the Kaufman Assessment Battery for Children; behavioural development was evaluated with the Child Behavior Checklist. The effect of PGS was analysed with a mixed-effects model. Main results and the role of chance Based on the intention to treat analysis, neurodevelopmental outcome of PGS children was similar to that of controls. However, additional analyses indicated that PGS affected neurodevelopmental outcome of twins in a different way than that of singletons. The fluency score of singletons born following PGS was similar to that of control singletons (mean values, 95% confidence intervals (CIs): 12.2 [11.5 to 12.8] and 12.2 [11.6 to 12.8] respectively, P = 0.977) that was also true for the other neurodevelopmental parameters. The fluency score of PGS twins was significantly lower than that of control twins (mean values, 95% CIs: 10.6 [9.8 to 11.3] and 12.3 [11.5 to 13.1] respectively, P = 0.001); the same was true for the NOS. In addition, PGS in twins was associated with a higher sequential intelligence quotient score. On the other hand, other neurodevelopmental parameters were similar for PGS twins and control twins. post hoc sample size calculation for the primary outcome parameter, the fluency score, indicated that the study groups, including the subgroups of singletons and twins, were adequately powered. Limitations, reasons for caution We assessed singletons and twins who contributed to the generalizability of the study. A limitation of our study is the relative small size of our study groups and the selective drop-out in both groups (drop-outs PGS group: higher gestational age; control group: less well-educated parents). These preclude the conclusion that PGS per

7


103

se is not associated with neurodevelopmental, cognitive and behavioural problems in singletons and the conclusion that PGS is associated with altered neurodevelopmental outcome in twins. Wider implications of the findings The need for careful long-term monitoring of children born following embryo biopsy remains, as it is still applied in the form of PGD and it is still unknown whether embryo biopsy affects long-term neurodevelopmental outcome.

7

104

INTRODUCTION Preimplantation genetic screening (PGS) was introduced to enhance efficiency in assisted conception. In PGS, a blastomere is aspirated from a cleavage stage embryo via an opening in the zona pellucida that is created by enzymatic digestion or with laser. Next, the copy numbers of several sets of chromosomes are identified by means of fluorescent in situ hybridization and only euploid embryos are transferred to the uterus. In theory, PGS results in improved implantation and increased ongoing pregnancy and live birth rates, as it is assumed that about half of all embryos obtained through IVF are aneuploid.21 However, a meta-analysis, performed over nine well-designed randomized controlled trials on PGS outcome, reported no evidence for a beneficial effect of PGS on live birth rate.23 Instead, PGS significantly reduced the live birth rate for women of advanced maternal age.23

Despite the invasiveness of the embryo biopsy inherent to PGS, few studies addressed development and growth of children born following PGS. Two research groups reported on developmental outcome of mixed groups of children born following embryo biopsy, i.e. born following PGS or PGD. One series of studies reported on mental, motor, socio-emotional and language development in 2-year-old PGD/PGS-children. The studies demonstrated similar outcome in singletons born following PGD/PGS and singletons born following ICSI or naturally conceived singletons.228 Similar results were found for twins.228 Another research group reported on health and developmental outcome measured with the Griffiths scale in children born following PGD/PGS up to the age of 4 years. Health and developmental outcome of the PGD/PGS children was similar to that of controls, with the exception of locomotor development, where PGD/PGS children scored significantly lower than controls.229 The studies did, however, not differentiate between PGD and PGS, although the indication to perform one or another is rather different. In PGD, blastomeres are analysed in couples with a high risk of a genetic disorder. In general, couples who are considered for PGD treatment do not have fertility problems, whereas couples who are considered for PGS treatment do. Underlying subfertility problems may contribute to the occurrence of neurodevelopmental problems in PGS offspring, as subfertility is known to be associated with more obstetrical events and more perinatal adversities.59,230

Previously, our group reported that neurodevelopmental outcome up until 2 years of age in children born following IVF with PGS was largely similar to that of children born following IVF without PGS.141,142 However, application of the detailed neurological optimality score (NOS) revealed that children born after IVF with PGS had a somewhat less optimal neurological condition at 2 years than children born after IVF without PGS. This may suggest a less favourable neurological development in children born following PGS.

The literature summarized above indicates that our knowledge on the long-term consequences of embryo biopsy on child development is unclear. We, therefore, decided to reassess the children born following IVF with and without PGS who participated in the follow-up study at 2 years.142 Reassessment consisted of the evaluation of neurodevelopmental, cognitive and behavioural outcome at the age of 4 years. The study is

7


105

part of a prospective, assessor-blinded, follow-up study of children born to women who participated in multicentre RCT on IVF with or without PGS.22

Primary outcome was the neurological condition of the child expressed in terms of fluency of motor behaviour. This is a sensitive measure to detect minor changes in neuromotor development, as fluency of movements is easily affected by subtle dysfunction of the brain. We therefore expected that potential differences between the two groups would be most clearly expressed in the fluency score. Secondary outcome parameters were the NOS, adverse neurological outcome – defined as the occurrence of complex minor neurological dysfunction (MND) or a neurological syndrome – and cognitive function and behaviour. We hypothesized that neurodevelopmental outcome, in particular fluency of movements, of children born following IVF with PGS is less optimal than that of children born after IVF without PGS, as PGS involves embryo biopsy. Specific attention was paid to the effect of PGS on singletons and twins, as Liebaers et al. indicated that embryo biopsy in combination with multiple pregnancy was associated with increased perinatal morbidity and mortality, whereas a similar increased morbidity and mortality was absent after embryo biopsy in singleton pregnancies.231

MATERIAL AND METHODS Participants For the present study all children born to women participating in a double blind, RCT on the efficiency of PGS to improve ongoing pregnancy rates after IVF were eligible (ISRCTN76355836).22 Inclusion criteria for participating women were: 1) age 35 to 41 years, 2) no previously failed IVF-cycles and 3) no objections against a possible double embryo transfer (DET). Randomization of women was performed centrally, using a computer program, with minimization for age (35-37 or 38-41 years) and reproductive technology (IVF or ICSI), and with stratification according to study centre [Academic Medical Center (AMC), Amsterdam and University Medical Center Groningen (UMCG), Groningen] prior to the start of the IVF procedures. Group status was revealed to participating parents after 12 weeks of gestation. Details on study design and information concerning IVF treatment procedures and the follow-up program have been reported previously.22,142

Prior to inclusion and randomization, couples were informed on the neurodevelopmental follow-up evaluation as part of the PGS trial. The Dutch Central Committee of Research Involving Human Subjects and the Medical Ethics Committees of the local hospitals approved the protocol of the follow-up study. The children’s parents provided written informed consent for participation of their child(ren) in the follow-up study, including the present follow-up.

7

106

Settings Demographic information, for example parity, gestational age, birthweight, neonatal intensive-care unit admission, parental age and educational level, was collected on standardized charts at the first follow-up assessment at 2 weeks.141

The assessment at 4 years was carried out by trained assessors supervised by a neurodevelopmental expert (M.H-A). The assessors were blind to prenatal and perinatal history, including the mode of conception. All assessments were carried out between October 2009 and November 2010. The children who were conceived at the UMCG were assessed at the Institute of Developmental Neurology at the UMCG, and the children who were conceived at the AMC were assessed at home.

Measures Neurological development Around the time of their fourth birthday, children were assessed with the standardized neurological assessment according to Hempel.182 The Hempel assessment was developed to detect MND at preschool age. Neurological signs are organized in five domains of function: fine motor function, gross motor function, posture and muscle tone, reflexes and visuomotor function (i.e. function of the visual system and eye movements). According to specified criteria each domain is classified as typical or deviant.155

Children were classified as neurologically normal, simple MND, complex MND or neurologically abnormal. A child is classified as having simple MND when one functional domain (except for the domain reflexes) is scored as deviant. Simple MND is regarded as a non-optimal, yet normal form of brain function.90 A child is classified as having complex MND when more than one domain is scored as deviant. Complex MND represents the clinically relevant form of MND, as it is associated with prenatal and perinatal adversities and with learning and behavioural disorders.90,157 Neurologically abnormal implies the presence of a distinct neurological syndrome such as cerebral palsy. Neurologically normal implies the absence of neurological dysfunction and is scored when no domains are deviant or only the domain of reflexes is deviant. We considered complex MND and neurologically abnormal as adverse neurological outcome.

Additionally, the outcome of the Hempel assessment was expressed in a NOS. The NOS consists of 56 items (range 0-56), for which an optimal condition is defined. The total score results from the sum of the items that fulfil the criteria for optimality. Higher scores represent better performance. It is important to realize that there is a conceptual difference between normality and optimality because the range for optimal behaviour is narrower than that for normal behaviour.160 This implies that the NOS is sensitive to minor changes, including those lying within the normal range. The fluency score (range 0-15) is a subscore of the NOS that evaluates the fluency of motor behaviour. The fluency score is the most sensitive measure to detect minimal changes in neuromotor development because subtle dysfunction of the nervous system is most easily expressed in a reduction in the fluency of

7


107

movements.161,162 The inter-rater reliability of the Hempel assessment is satisfactory [reliability per item: κ = 0.62-1.00 (mean 0.93)] and its construct validity is good.(155, 163-165)

Cognitive development Cognitive development was evaluated by means of the Kaufman Assessment Battery for Children, second edition (K-ABC-II).166 The K-ABC-II is an individually administered standardized clinical instrument that measures cognitive and processing abilities in children and adolescents (3 – 18 years). Cognitive and processing abilities are expressed in a total intelligence quotient (IQ) score (the so-called Fluid-Crystallized Index) and four IQ scale scores: 1) a sequential processing IQ that reflects the short-term memory of a child, 2) a simultaneous processing IQ, representing the spatial aptitude of a child, 3) a learning ability IQ, representing the long-term memory capacity of a child and 4) a knowledge IQ. All raw scores are normalized into global scores (mean: 100; standard deviation [SD]: 15). The K-ABC-II has been used for children with different social backgrounds or ethnic differences without critical effects on test-scores. Reliability and validity of the K-ABC-II are good.166 American norms were applied because Dutch norms are lacking.

Behavioural development Behavioural development was evaluated by means of the Dutch version of the Child Behavior Checklist (CBCL).167,168 The CBCL is a parental questionnaire designed to identify emotional and behavioural problems in children aged 1½ to 5 years. Questions on the CBCL are grouped into the following problem scales; emotionally reactive, anxious/depressed, somatic complaints, withdrawn, sleep problems, attention problems and aggressive behaviour. The first four of these scales together form the internalizing scale and the latter two form the externalizing scale. The sum of all questions determines the total problem scale. Raw scores are normalized into T-scores (mean: 50; SD: 10). Higher scores represent more problematic behaviour. Reliability and validity of the CBCL are good.167 Statistical analysis Fisher’s exact tests and Mann-Whitney U-tests were applied to compare demographic characteristics between the PGS group and the control group at parent level and at child level separately for singletons and twins.

A linear mixed-effects model, with a random effect for mother to model the possible correlation between twins, was performed to analyse the potential effect of PGS for twins and singletons separately on the primary outcome measure fluency score and the secondary outcome measures (NOS, the various IQ scores and the three CBCL scores). A type III t-test with Satterthwaite’s degrees of freedom173 was used for the specific contrasts. The effect of PGS on the additional secondary outcomes MND and adverse neurological outcome was estimated for singletons and twins separately using generalized estimating equations with a logit link function and a binomial distribution. The Wald test was applied

7

108

for the specific contrasts. For the numerical outcome measures, all analyses were adjusted for centre and assessor, whereas for the binary outcome measures (MND, adverse neurological outcome), all analyses were adjusted for centre only.

FIGURE I. Flow chart on eligibility and participation of children and parents in the PGS RCT and follow-up study.

7


109

A post hoc power analysis was conducted using a non-central t-test to estimate the minimal detectable effect size for singletons and twins separately or together, depending on the outcome of the test. The type I error rate and power were selected at 5% and 80% respectively.

All analyses were performed according to the ‘intention to treat’ (ITT) principle in which the four children born to couples who conceived naturally and the one child that was conceived via intra-uterine insemination (IUI) were taken into account. To explore potential effects of natural conception and IUI, we performed sensitivity analyses in which we repeated all analyses with the exclusion of the four naturally conceived children and the one IUI child. The sensitivity analysis also included an exploration of a potential interaction effect of ICSI with IVF for singletons and twins. Finally, we adjusted for gestational age in the primary analyses to investigate whether or not the effect of PGS was mediated by gestational age, as it is known that gestational age is an important predictor for neurological outcome.

All analyses were performed with the Statistical Analysis System software version 9.2 or the Predictive Analytics SoftWare Statistics, version 18. In all analyses probability values of 5% or less were considered significant.

RESULTS

Participation and demographic characteristics Between May 2003 and November 2005, 408 women were included in the randomized trial. The trial resulted in 52 ongoing PGS pregnancies and 74 control pregnancies22 of which respectively 49 (39 singletons, 10 sets of twins) and 71 (57 singletons, 14 sets of twins) pregnancies resulted in live births (Figure I). A set of twins died immediately postpartum due to immature birth. Four couples could not be invited for follow-up due to the withdrawal of informed consent during treatments. The reason for withdrawal was in most cases the stress caused by the blinding of couples for treatment allocation. Eventually, 47 PGS couples and 68 control couples were eligible for follow-up evaluation. At 4 years of age, 40 (85%) PGS couples with 49 children (31 singletons, 9 sets of twins) and 53 (78%) control couples with 64 children (42 singletons, 11 sets of twins) participated in the study (attrition rate of 18% with regard to baseline, Figure I). Overall, background factors of non-participants were similar to those of participants, with the exception of selective drop-out of less well-educated control parents (mothers: P = 0.006; fathers: P = 0.040), and PGS children with a higher gestational age (P = 0.040).

Demographic characteristics of parents and children of both study groups are shown in Table I. At parent level, demographic characteristics of the groups were similar, except for Caesarean section: PGS children were less often delivered by Caesarean section (P = 0.039). At child level, all demographic characteristics of the groups were similar, also in the subgroups of singletons and twins, except for gestational age: children of the PGS group

7

110

all ch

ildre

nsin

gleto

nstw

ins

all ch

ildre

nsin

gleto

nstw

ins

n = 4

0n

= 49

n = 3

1n

= 18

n = 5

3n

= 64

n = 4

2n

= 22

Pare

ntal c

hara

cteris

tics

Mater

nal a

ge at

conc

eptio

n in y

ears,

med

ian (r

ange

)37

.3 (3

5.2 -

40.8)

37.9

(35.3

- 41

.3)Ed

ucati

on le

vel m

other

(high

a ), n (

%)

21 (5

3)31

(59)

Educ

ation

leve

l fathe

r (hig

ha ), n (

%)

22 (5

5)27

(51)

Time t

o pre

gnan

cy (y

ears)

e , med

ian (r

ange

)3.9

(1.1

- 12.4

)4.1

(0.7

- 15.9

)Sm

oking

durin

g pre

gnan

cye , n

(%)

3 (9)

1 (2)

Alcoh

ol co

nsum

ption

durin

g pre

gnan

cye , n

(%)

1 (3)

1 (2)

Conc

eptio

n meth

odIV

F pe

rform

ed, n

(%)

24 (6

0)29

(55)

ICSI

perfo

rmed

, n (%

)13

(33)

22 (4

2)IU

I per

forme

db , n (%

)1 (

3)0

Natur

alc , n (%

)2 (

5)2 (

4)PG

S tre

atmen

t cen

tre (U

MCG)

, n (%

)22

(55)

28 (5

3)Ch

ild ch

arac

terist

icsTw

ins, n

(%)

18 (3

7)22

(34)

Male

gend

er, n

(%)

26 (5

3)17

(59)

9 (50

)34

(53)

19 (4

5)15

(68)

Corre

cted a

ge at

exam

inatio

n at 4

year

s of a

ge (m

onths

), me

dian (

rang

e)50

.9 (4

7.8 -

67.9)

50.9

(47.8

- 67

.9)51

.1 (4

8.4 -

59.8)

50.2

(47.8

- 70

.2)50

.3 (4

7.8 -

70.2)

49.5

(48.4

- 60

.8)

Birth

char

acter

istics

First

born

, n (%

)24

(60)

12 (3

9)8 (

44)

29 (5

5)12

(29)

6 (27

)Ge

statio

nal a

ge (w

eeks

), me

dian (

rang

e)

39.4

(32.0

- 42

.0)*

39.7

(36.9

- 42

.0)**

37.1

(32.0

- 39

.3)38

.7 (3

0.7 -

41.6)

*39

.1 (3

0.7 -

41.6)

**36

.7 (3

3.0 -

38.7)

Prete

rm bi

rth (<

37 w

eeks

), n (

%)

5 (13

)1 (

3)8 (

44)

9 (17

)2 (

5)14

(64)

Birth

weig

ht (g

rams

), me

an (S

D)

3219

(711

)35

72 (5

46)

2611

(532

)30

73 (6

57)

3351

(566

)25

42 (4

69)

Low

birth

weigh

t (<

2500

gram

), n (

%)

9 (18

)2 (

7%)

7 (39

)12

(19)

1 (2)

11 (5

0)Sm

all fo

r ges

tation

al ag

ed , n (%

)5 (

10)

3 (10

)2 (

11)

3 (5)

0 (0)

3 (14

)Ca

esar

ean s

ectio

ne , n (%

)7 (

18)*

20 (3

8)*

Neon

atal c

hara

cteris

tics

Apga

r sco

re 5

min <

7,e n

(%)

0 (0)

0 (0)

0 (0)

2 (3)

1 (2)

1 (5)

Neon

atal in

tensiv

e car

e adm

ission

, n (%

)9 (

18)

1 (3)

8 (44

)6 (

9)2 (

5)4 (

18)

Note:

Fish

er’s

exac

t tests

, Stud

ent's

t-tes

t and

Man

n-W

hitne

y U-te

sts w

ere a

pplie

d to c

ompa

re de

mogr

aphic

char

acter

istics

betw

een t

he P

GS gr

oup a

nd th

e con

trol g

roup

at pa

rent

level

and a

t chil

d lev

el se

para

tely f

or si

nglet

ons a

nd tw

ins. *

P <

0.05

, ** P

< 0.

01.

PGS

= pr

eimpla

ntatio

n gen

etic s

creen

ing; IV

F =

in vit

ro fe

rtiliza

tion;

ICSI

= in

tracy

toplas

mic s

perm

injec

tion;

IUI =

intra

uterin

e ins

emina

tion;

UMCG

= U

niver

sity M

edica

l Cen

ter G

ronin

gen.

a Univ

ersit

y edu

catio

n or v

ocati

onal

colle

ges.

b Trea

tmen

t was

conv

erted

to IU

I in ca

se of

poor

follic

le gr

owth.

c Data

were

analy

sed a

ccor

ding t

o inte

ntion

to tr

eat.

d Birth

weigh

t for g

estat

ional

age i

s < -2

stan

dard

devia

tions

comp

ared

with

the D

utch r

efere

nce p

opula

tion (

Dutch

refer

ende

table

s, pe

rinata

l Reg

istra

tion N

ether

lands

).e Mi

ssing

data

in tw

o gro

ups:

time t

o pre

gnan

cy n=

1, sm

oking

durin

g pre

gnan

cy n=

9, alc

ohol

cons

umpti

on du

ring p

regn

ancy

n=12

, Cae

sare

an se

ction

n=1,

Apga

r sco

re 5

min <

7 n=

8.

Char

acter

istics

TABL

E I. C

hara

cter

istics

of p

aren

ts an

d in

fant

s in

the P

GS fo

llow-

up st

udy.

Cont

rol c

hild

ren

Child

ren

born

after

PGS

Cont

rol c

oupl

esCo

uples

with

≥ 1

live b

irth

after

PGS

7


111

had a higher gestational age at birth (P = 0.027) than children of the control group. In the subgroups, PGS singletons had a higher gestational age at birth (P = 0.005) than controls; gestational age of twins was similar.

The mixed-effects model indicated that PGS affected neurodevelopmental outcome of twins in a different way than that of singletons. Therefore, the results on outcome are presented for singletons and twins separately.

Neurological development at 4 years Note that the results on neurological outcome in clinical terms of adverse neurological outcome are presented first. Thereafter, the results of the primary outcome parameter, the fluency score, and those of the NOS are presented.

Based on the ITT analysis, neurological outcome of PGS children was similar to that of controls. However, the mixed-effects models indicated that neurological outcome differed for singletons and twins (Table II). The rate of adverse neurological outcome in PGS children and controls was similar in singletons (PGS: n=3 (10%); controls: n=10 (24%), P = 0.145) and twins (PGS: n=6 (33%); controls: n=3 (14%), P = 0.116). However, one member of a PGS twin was diagnosed with cerebral palsy, whereas no members of control twins had cerebral palsy. The fluency score of singletons born following PGS was similar to that of control singletons (P = 0.977). However, PGS twins had significantly lower fluency scores than control twins [mean values, 95% confidence intervals (CIs): 10.6 (9.8 to 11.3) and 12.3 (11.5 to 13.1) respectively, P = 0.001]. Similarly, the NOS did not differ between PGS singletons and controls (P = 0.548) but PGS twins had a significantly lower NOS than control twins [mean values, 95% CIs: 44.8 (42.8 to 46.9) and 48.8 (46.8 to 50.9) respectively, P = 0.005].

After adjusting for gestational age in the analyses concerning the fluency score and the NOS, the differences between the groups remained statistically significant (P = 0.001 and P = 0.005, respectively). Exclusion of the PGS child with cerebral palsy did not alter the conclusions about the group differences in fluency score and NOS. Also the exclusion of the four naturally conceived children and the one IUI child did not alter the conclusions (data not presented). Moreover, the effect of PGS on the neurodevelopmental outcome measures did not differ for ICSI and IVF singletons and twins (data not presented).

Cognitive development at 4 years Total IQ scores of all children were in the normal range, except for one singleton of the control group who had a total IQ score of 82. Based on the ITT analysis, cognitive development of PGS children was similar to that of controls but again differences were found for the effect of PGS in singletons and twins (Table II). The total IQ scores of PGS singletons did not differ from those of control singletons (P = 0.666). Moreover, no differences were found in the four IQ scale scores between PGS singletons and controls, except for the learning IQ score that turned out to be significantly lower among PGS singletons when compared to control singletons (P = 0.049) (Table II). However, after adjusting for gestational age in the analyses, the difference disappeared (P = 0.066). The

7

112

total IQ scores of PGS twins did not differ from those of control twins (P = 0.104). Moreover, no differences were found in the four IQ scale scores between PGS twins and control twins, except for the sequential IQ score that turned out to be significantly higher among PGS twins compared to control twins (P = 0.027), also after adjusting for gestational age (P = 0.028). Exclusion of the PGS child with cerebral palsy did not alter the conclusions on group differences in cognitive outcomes. Also when the one IUI child and the four naturally conceived children were excluded, the conclusions did not change (data not presented). Moreover, the effect of PGS on the cognitive outcome measures did not differ for ICSI and IVF singletons and twins (data not presented).

Behavioural development at 4 years Based on the ITT analysis, behavioural development of PGS children was similar to that of controls. No differences between the groups were found in the total problem score in singletons (P = 0.357) and twins (P = 0. 983) or in internalizing and externalizing behaviour (Table II).

Again, the sensitivity analyses did not demonstrate alternative results when the one IUI child and the four naturally conceived children were excluded (data not presented). Moreover, we did not detect heterogeneity between the effect of PGS for ICSI and IVF singletons and twins on the behavioural outcome measures (data not presented). Also exclusion of the PGS child with cerebral palsy did not alter the conclusions about the group differences in behavioural outcomes.

Post hoc power analysis We performed a post hoc sample size calculation for the primary outcome parameter, the fluency score, for singletons and twins separately. The post hoc power analysis indicated that the current sample size could detect a difference of 1.0 and 1.4 on the fluency score with 80% power and 5% significance level for singletons and twins respectively. This means that our study groups are adequately powered for the primary outcome parameter.

Also, our singleton and twin groups are sufficiently powered for the NOS with a minimal detectable effect size of 2.6 and 3.8 respectively. However, no relevant clinical effect sizes could be detected with the current sample size for the outcome adverse neurological outcome (odds ratios of more than 7.3 and 8.2 for singletons and twins, respectively). For the cognitive outcomes, an effect size of less than seven points with 80% or more power was detectable for singletons, with the exception of the simultaneous IQ score and the knowledge IQ score, where the power was equal to 75% and 63%, respectively. For twins, an effect size of less than 11 points was detectable with more than 80% power for the learning IQ score and the total IQ score; for the simultaneous IQ score, the knowledge IQ score and the total IQ score the power was determined at 75%, 70% and 57%, respectively. For the behavioural outcomes the study groups were adequately powered.

7


113

Neur

olog

ical d

evelo

pmen

t

PGS

sing

leto

ns (n

= 30

)Co

ntro

l sin

glet

ons

(n =

41)

P -

valu

ePG

S tw

ins

(n =

18)

Cont

rol t

wins

(n =

22)

P -

valu

eP

- va

lue

Inte

ract

ion

effe

ct

sing

leto

n*tw

ins

Fluen

cy sc

ore,

mean

[CI]

12.2

[11.5

; 12.8

]12

.2 [11

.6 ; 1

2.8]

0.977

10.6

[9.8 ;

11.3]

12.3

[11.5

; 13.1

]0.0

010.0

05Ne

urolo

gical

optim

ality

scor

e (NO

S), m

ean [

CI]

49.3

[47.6

; 50.9

]48

.7 [47

.1 ; 5

0.3]

0.548

44.8

[42.8

; 46.9

]48

.8 [46

.8 ; 5

0.9]

0.005

0.007

PGS

sing

leto

ns (n

= 30

)Co

ntro

l sin

glet

ons

(n =

42)

PGS

twin

s (n

= 18

)Co

ntro

l twi

ns (n

= 22

)P

- va

lue

Inte

ract

ion

effe

ct

sing

leto

n*tw

ins

Mino

r neu

rolog

ical d

ysfun

ction

(MND

)a , n (%

)8 (

27)

16 (3

8)0.3

126 (

33)

8 (36

)0.8

590.6

15Ad

verse

neur

ologic

al ou

tcome

b , n (%

)3 (

10)

10 (2

4)0.1

456 (

33)

3 (14

)0.1

160.0

36

Cogn

itive

dev

elopm

ent

PGS

sing

leto

ns (n

= 30

)Co

ntro

l sin

glet

ons

(n =

42)

P -

valu

ePG

S tw

ins

(n =

16)

Cont

rol t

wins

(n =

22)

P -

valu

eP

- va

lue

Inte

ract

ion

effe

ct

sing

leto

n*tw

ins

Total

IQ, m

ean [

CI]

113.4

[109

.7 ; 1

17.4]

114.4

[110

.7 ; 1

18.1]

0.666

109.8

[ 10

4.2 ;

115.5

]10

4.1 [9

8.8 ;

109.3

]0.1

040.1

05Se

quen

tial IQ

, mea

n [CI

]98

.0 [93

.7 ; 1

02.3]

100.7

[96.6

; 10

4.9]

0.259

103.5

[96.9

; 11

0.2]

94.2

[88.1

; 100

.3]0.0

270.0

14 Le

arnin

g IQ,

mea

n [CI

]98

.7 [94

.3 ; 1

03.1]

103.5

[99.2

; 10

7.8]

0.049

96.3

[90.7

; 101

.9]95

.1 [89

.7 ; 1

00.4]

0.710

0.143

Simult

aneo

us IQ

, mea

n [CI

]12

3.2 [1

18.4

; 128

.0]12

1.2 [1

16.6

; 125

.8]0.4

4711

6.0 [1

09.0

; 123

.1]11

3.2 [1

06.7

; 119

.7]0.5

190.8

76Kn

owled

ge IQ

, mea

n [CI

]11

8.6 [1

13.2

; 124

.0]11

6.0 [1

10.8

; 121

.2]0.3

8911

1.6 [1

03.5

; 119

.8]10

6.8 [9

9.3 ;

114.3

]0.3

410.7

06

Beha

viour

al de

velo

pmen

t

PGS

sing

leto

ns (n

= 30

)Co

ntro

l sin

glet

ons

(n =

41)

P -

valu

ePG

S tw

ins

(n =

18)

Cont

rol t

wins

(n =

22)

P -

valu

eP

- va

lue

Inte

ract

ion

effe

ct

sing

leto

n*tw

ins

CBCL

total

scor

e, me

an [C

I]45

.7 [41

.8 ; 4

9.6]

47.7

[44.0

; 51.5

]0.3

5746

.1 [40

.6 ; 5

1.5]

46.1

[40.8

; 51.5

]0.9

830.6

41 C

BCL i

ntern

alizin

g sco

re, m

ean [

CI]

47.9

[41.7

; 52.1

]49

.0 [44

.9 ; 5

3.1]

0.630

47.5

[41.8

; 53.2

]47

.4 [41

.7 ; 5

3.1]

0.981

0.779

CBCL

exter

naliz

ing sc

ore,

mean

[CI]

47.5

[44.0

; 51.1

]48

.4 [44

.9 ; 5

1.8]

0.684

47.8

[42.8

; 52.8

]47

.3 [42

.4 ; 5

2.3]

0.902

0.750

TABL

E II.

Mixe

d-ef

fect

mod

el an

alyse

s on

neur

olog

ical, c

ogni

tive a

nd b

ehav

iour

al ou

tcom

e with

or w

ithou

t pre

impl

anta

tion

gene

tic sc

reen

ing.

a MND

= sim

ple an

d com

plex m

inor n

euro

logica

l dys

functi

on.

b Adv

erse

neur

ologic

al ou

tcome

= co

mplex

MND

or w

orse

(suc

h as c

ereb

ral p

alsy).

IQ =

intel

ligen

ce qu

otien

t, CBC

L = C

hild B

ehav

iour C

heck

list, C

I = co

nfide

nce i

nterva

l.No

te tha

t the n

umer

ical o

utcom

es ar

e adju

sted f

or ce

ntre (

UMCG

, Gro

ninge

n, Gr

oning

en or

AMC

, Ams

terda

m) an

d ass

esso

r and

that

the bi

nary

outco

me m

easu

res a

re ad

justed

for c

entre

only.

7

114

Overall, the post hoc analyses indicate that the study sample sizes are large enough to detect relevant differences between PGS and controls singletons and twins separately for the primary outcome parameter and for most secondary outcome parameters.

DISCUSSION The present study did not demonstrate statistically significant differences in neurological, cognitive and behavioural outcome at 4 years in singletons born following IVF with PGS – involving embryo biopsy – and those born following IVF without PGS. However, we demonstrated statistically significant differences between twins born following IVF with PGS and twins born following IVF without PGS. Our findings on singletons are similar to those of the two other groups that reported on developmental outcome of singletons born after embryo biopsy.229,232,233 However, it should be noted that these studies did not differentiate between PGD and PGS, although the indication to perform one or another is rather different. Our finding that PGS affected developmental outcome of twins does not, however, correspond to studies of others.228,229 It matches, however, to some extent the study of Liebaers et al., who found higher rates of prematurity and low birthweight in PGD/PGS multiples than in ICSI multiples and more perinatal deaths in post PGD/PGS multiple pregnancies than in post ICSI multiple pregnancies, whereas outcomes were similar for PGD/PGS and ICSI singletons.231 Yet, in a larger analysis by Desmyttere et al., including the analyses reported previously by Liebaers et al., no differences in prematurity, low birthweight and perinatal deaths were found between PGD/PGS multiples in comparison with ICSI multiples.234,235 This means that the effects of PGD/PGS are inconsistent. It is also good to realize that our results do not necessarily suggest an overall adverse effect of PGS on outcome of twins, as PGS was associated with a negative effect on neuromotor condition and a positive one on sequential processing, an indicator of short-term memory. Our results rather suggest that PGS affects neurodevelopmental outcome of twins in a different way than that of singletons.

We previously reported on outcome of the same group of children exclusively born after PGS. At 2 years of age, we found that children born after IVF with PGS had, on average, an approximately two points lower NOS than controls.142 At 4 years of age we found a similar effect in PGS twins, who scored approximately two points lower on the fluency score and approximately four points lower on the NOS when compared to the controls. However, such an effect was not present in singletons at 4 years because the mean difference reduced to less than one point. At the 2 years of follow-up singletons and twins were analysed together, which means that we do not know whether the effect of PGS at that time also could be attributed to the effect on twins. The current findings, however, indicate that PGS may affect neurological condition of twins but not that of singletons. In addition, the appearance of a clearer effect of PGS at 4 years may be related to the characteristics of the developing brain. It is well known that with increasing age, children may grow into a

7


115

deficit, i.e. that neurological dysfunctions first emerge when the neural circuitries subserving specific functions become functionally active.90

Strengths and limitations To our knowledge, this is the first follow-up study of a RCT on PGS in which children were followed as long as 4 years. The random assignment of parents resulted in a strong resemblance of the two groups with respect to most background variables so that comparisons between both groups primarily reflect the effect of PGS. Additional strengths with regard to study design are the blinding of the assessors to the mode of conception and the prospective design of the study. Parents were invited to participate before pregnancy so that potential selection bias based on the child’s development or health was reduced. Another strength of the present study is the examination of both singletons and twins, as it is known that being a member of a multiple is associated with an increased risk of developmental problems, irrespective of assisted conception.179 Moreover, the examination of both singletons and twins contributes to the generalizability of the study. It may be argued that another strength is that we included in both PGS and control groups children born after IVF and ICSI. IVF and ICSI are both applied whenever PGS is indicated or not. This contrasts with PGD treatment, in which IVF with ICSI is virtually always used. However, it could be argued that ICSI substantially differs from IVF alone. We therefore applied not only minimization for reproductive technique (IVF or ICSI) during enrolment but also a sensitivity analysis, of which the results indicated that the presence or absence of ICSI did not modify a potential effect of PGS.

A major strength is the application of sensitive and age-specific measurements to assess neurological condition. The strength of the Hempel assessment is illustrated by the study of Bouwstra et al., in which a negative effect of neonatal trans-fatty acid status on neurodevelopmental outcome was demonstrated with the assessment according to Hempel but not with the Bayley’s Scale of Infant Development.164 Although subtle differences in neurological outcome, such as a few points reduction in the NOS or fluency score, may not have clinical relevance for individuals, minor differences in neurodevelopmental outcome may affect a particular subgroup of society, such as subfertile people, a group that is steadily expanding.236

The major limitation of the present study is the relatively small sample size. Power calculation was based on the number of women needed to detect a certain increase in the cumulative ongoing pregnancy rate.22 The unforeseen effect of lower pregnancy rates after PGS further reduced the number of children available for follow-up. We performed a post hoc sample size calculation for the primary outcome parameter, the fluency score, which indicated that our study groups were adequately powered also for the subgroups of singletons and twins. For several other secondary outcome measures the minimal detectable effect sizes were still acceptable, and only for the outcome parameter adverse neurological dysfunction was the study too small. This means that in general our study groups are adequately powered because relevant differences could have been detected.

7

116

However, it should be noted that there was a selective loss of less well-educated parents in the control group and a selective loss of children with higher gestational age in the PGS group. Perhaps parents of infants with a lower gestational age are more concerned about their child’s development and therefore more willing to participate in a detailed neurodevelopmental assessment. Presumably, the selective loss of less well-educated parents in the control group does not adequately explain the different neurodevelopmental outcome of PGS twins when compared to control twins, as the selective loss was present in twins and singletons, where the selective loss in the singletons was not associated with developmental differences between PGS children and controls. Moreover, we explored the effects of confounding variables on our outcome measures in an explorative analysis to make sure that the effects found were not mediated by these confounders. This analysis revealed among others no statistically significant effects of parental educational level or gestational age.

The groups, especially the twin groups, were relatively small, implying that the distribution of scores should be taken into account. For the NOS and fluency, Pearson's residuals did not demonstrate a violation of the assumption of normality, but one twin with cerebral palsy had a score of more than 3 SDs away from the predicted value. This twin can be defined as an outlier in our data. However, excluding this twin from the analyses did not change the conclusions on the significance of the effect of PGS on twins for the NOS and fluency.

The participating children of both groups turned out to have a relatively high IQ. An explanation for this finding may be the relatively large proportion of highly educated parents participating in our study, as it is well known that cognitive abilities of children are positively associated with their parents education level.237 This means that our findings have to be interpreted with caution.

The application of PGS in women of advanced age is debated.23 The European Society of Human Reproduction and Embryology PGD Consortium recommends to apply PGS only in the context of properly constructed trials.24 Even though the use of PGS currently seems to be decreasing,238 embryo biopsy is still applied, for example, in the form of PGD.236 Another issue of concern is the application of DET in IVF with embryo biopsy. DET in IVF with embryo biopsy may entail an accumulation of several risks for offspring, as DET per se increases the risk of twin pregnancy,239,240 and perinatal mortality241 when compared to single embryo transfer in IVF.

In conclusion, the results of the present study suggest that neurological, cognitive and behavioural outcome at 4 years in singletons born following IVF with PGS is similar to that of singletons born after IVF without PGS. In contrast, PGS does seem to affect neurodevelopmental outcome of 4-year-old twins. PGS in twins was associated with a difference in brain function at 4 years, i.e. with a negative effect on neuromotor condition and a positive one on sequential processing. This may point to the possibility that the embryo biopsy inherent to PGS is associated with differences in brain function at a later age.

7


117

Hence, the need for careful monitoring of children born following embryo biopsy remains, for it is not yet known whether there are any late consequences of embryo biopsy on neurodevelopmental outcome.

ACKNOWLEDGEMENTS

We thank participated parents and children for their cooperation and enthusiasm during the assessments; Maaike Haadsma, M.D., Ph.D., for her help in including the participants; Rosan Aapkes, B.Sc., Elise Bennik, M.Sc., Hanneke Geut, M.D., Marjolein Jongbloed-Pereboom, M.Sc. and Bertine de Vries, M.Sc., for support in collecting the data; Michiel Schrier, M.Sc., Linze Dijkstra B.Sc., and Loes de Weerd for technical assistance.

7

118

General discussion

119

CHAPTER 8 General discussion

120

This thesis evaluates the effects related to assisted conception on neurodevelopmental outcome in preschool-aged children. Standardized, age-specific and sensitive tools were applied to study neurodevelopmental outcome in the participants of two parallel running projects, the Groningen ART (assisted reproductive techniques) cohort study and the PGS (preimplantation genetic screening) study. In general, this thesis demonstrated that ART-related aspects such as ovarian hyperstimulation, the in vitro laboratory procedures, or a combination of both were not associated with worse neurodevelopmental outcome – in terms of movement variation, neuromotor function, cognition and behaviour – in singleton children aged 4 months to 4 years. Additionally, a history of subfertility per se and the underlying cause of subfertility were not associated with worse neurodevelopmental outcome at age 2 and 4. However, increased time to pregnancy (TTP) was associated with an increased risk for minor neurological dysfunction (MND) at 2 years and an increased risk for the complex form of MND at 4 years. Moreover, direct negative effects were found between both the severity of subfertility in terms of TTP and cognition and the presence of subfertility and behaviour at age 4.

In this thesis it was also observed that neurodevelopmental outcome – in terms of neuromotor function, cognition and behaviour – in singletons born following in vitro fertilization (IVF) with PGS is similar to that of singletons born after IVF without PGS at age 4. However, PGS may be associated with altered neurodevelopment in twins.

This chapter starts with a discussion of the main findings of the studies conducted. First, the main findings of the Groningen ART cohort study will be discussed, in terms of ART-related and subfertility-related aspects related to child neurodevelopmental outcome. Hereafter, the chapter focusses on the implications of ART and subfertility for daily practice and the consequences of minor neurological signs. Next, the main findings of the PGS study, will be discussed, including the clinical implications of embryo biopsy. The chapter continues with an overview of methodological considerations per project, some suggestions for future research and ends with the concluding remarks.

ART aspects related to child neurodevelopmental outcome This thesis concludes that rather the severity of subfertility than its presence or underlying cause or ART-related aspects such as ovarian hyperstimulation and the in vitro procedures affect neurodevelopmental outcome in preschool-aged singleton children, born to subfertile couples. This finding corroborates previous findings within the Groningen ART cohort study: similar neurodevelopmental outcome between the three ART study groups were found at 2 weeks and 3, 4, 10 and 18 months of age.135,140,242 Moreover, our observation that ART-related aspects do not negatively affect the child’s neurodevelopmental outcome correspond to the majority of follow-up studies.78,99-106 Does this mean that ART is entirely safe for the developing child? Although generally no significant differences were found between COH-IVF, MNC-IVF and Sub-NC children, a closer look at the results in this thesis learns that once a difference was found between group COH-IVF and MNC-IVF. In the period

8

General discussion

121

of 4 to 18 months of age, COH-IVF children had lower IMP variation scores than MNC-IVF children. An explanation for the finding that ART affects the IMP variation score rather than the prevalence of MND or the optimality scores may be the fact that variation in motor behaviour taps on other neural circuitries than traditional neurological signs.174 Variation in motor behaviour in particular reflects integrity of cortical connectivity.145,150 A similar specific effect of ovarian hyperstimulation was observed in another study that was part of the current Groningen ART cohort study in terms of cardiometabolic outcome. Seggers et al. and La Bastide-Van Gemert et al., demonstrated a direct negative effect of ovarian hyperstimulation on systolic blood pressure percentiles and subscapular skinfold thickness at age 4.217,243 A potentially adverse effect of ovarian hyperstimulation could be mediated by an altered oocyte quality, due to the production of multiple follicles by the ovaries, resulting in non-optimal oxygenation of the rapidly growing follicles and the bypass of the dominant follicle.244 Furthermore, both animal and clinical and studies suggest that supraphysiological estradiol levels may be toxic for the developing embryo by impairing its implantation potential.245,246 Ovarian hyperstimulation may also interfere with the intra-uterine environment.40 The alterations in hormonal levels may lead to disturbances in the endometrium, resulting in diminished endometrial receptivity, a non-optimal expression of endometrial growth and thus, a poorer implantation environment.41,42,247 The altered early environment could induce a non-optimal development of the oocyte and embryo. A study by Pelinck et al. supports the latter thought. The authors reported that COH-IVF children had a 134 grams lower birthweight compared to MNC-IVF, suggesting that ovarian hyperstimulation may be a causative factor rather than the in vitro procedure per se in the occurrence of low birthweight after conventional IVF.248

However, an argument against an unfavourable effect of ovarian hyperstimulation is the finding that the variation scores of children born following COH-IVF did not differ significantly from that of Sub-NC children. Moreover, for all of the other neurodevelopmental outcome parameters, no differences were found between COH-IVF children and Sub-NC children during the whole follow-up period from 4 months of age until 4 years of age. Many other studies also conclude that early neurodevelopmental outcome of children born after IVF – i.e. the commonly applied IVF with ovarian hyperstimulation – does not differ from children conceived naturally (NC).78,99-106

Subfertility aspects related to child neurodevelopmental outcome This thesis concludes that the presence of a history of subfertility per se or the underlying cause of subfertility do not seem to affect the neuromotor development of children at the age of 2 and 4 years. At 2 years, no differences in optimality scores or the prevalence of MND could be observed between the children born to the Groningen ART cohort and children born to a fertile reference group. Similar results were found for the optimality scores and prevalence of complex MND at 4 years. These are initially reassuring findings, however, the absence of a difference in outcome between children born to subfertile parents and those born to fertile parents may be explained by the composition of the

8

122

reference group. An argument for this hypothesis lies in the study of Middelburg et al. on neurodevelopmental outcome of the Groningen ART cohort children at 3 months.135 In this study, a significantly worse neurological outcome was found for the cohort children compared to compared to a reference group that consisted of a representative sample of the general population, as the infants had been assessed as a part of a general health check-up provided for all infants.135 At the follow-up at 2 and 4 years – when no significant neurological disadvantage was found for the subfertile group – the fertile reference group consisted of children whose parents volunteered for participation in the study. This may have caused selection bias. Perhaps, parents who are more concerned about the health and development of their child are more willing to participate in a neurodevelopmental outcome study.196 The idea of selection bias is supported by the relatively high proportion of children in the fertile reference group with complex MND (14%), a proportion that is higher than that reported in the general population (6-7%).183 A similar phenomenon was seen in the study by Knoester et al. That study reported a higher prevalence of MND in the control group of NC children compared to the general population.104 Given this potential selection bias, the slightly worse neurodevelopmental outcome the authors found in their 5 to 8-year-old ICSI children compared to their NC children may have actually been larger.

The composition of the fertile reference group may be an explanation for the absence of an effect of subfertility per se on child neuromotor outcome at 2 and 4 years. However, in the study on the ART treatment effects and subfertility effects on cognition and behaviour in 4-year-olds (Chapter 6), a direct negative causal effect was found between the presence of subfertility and behaviour, an effect that came to light with the inclusion of the fertile reference group. Also, a direct negative causal effect was found of the severity of subfertility, reflected by a longer TTP, and the child’s cognition. Both effects were confounded by maternal age at child conception and maternal educational level and cognition and behaviour were directly related to one another. This implies that suffering from subfertility per se, and especially from more severe subfertility – which by itself is associated with higher age and high educational level of the mother – negatively affects the child’s cognitive and behavioural outcome. A similar finding was demonstrated in Chapter 4 and 5, where it is concluded, that a longer TTP was associated with a less optimal neurological condition in preschool-aged children. The effect of TTP on neurodevelopmental outcome was investigated at age 2 and age 4. At 2 years of age, TTP was associated with a higher prevalence of MND, whereas at 4 years of age it was associated with a higher prevalence of complex MND. In the 4-year follow-up round, special attention was paid to potential sex-differences in outcome. The negative association found in the study was particularly true for girls. In general, boys have twice as often complex MND as girls,90,183 which may be related to the sex-specific differences and timing of the development and maturation of the nervous system and, with that, the sex-depending critical period for the occurrence of developmental events.197-200 However, in our subfertile study group a relatively high proportion of children, boys as well as girls, had complex MND (total group: 24%; boys: 21%; girls: 28%), a proportion that is considerably higher than in the

8

General discussion

123

general population (6-7%).183 One could wonder why a longer TTP was only associated with clinical neurological condition in terms of complex MND and not with the derived parameters fluency score and NOS. Clinical neurological condition and neurological optimality are highly correlated; however, a reduction in neurological optimality does not necessarily mean the presence of a neurological deviation, as the range for optimality is narrower than that for normality.160 Moreover, the classification of MND involves a threshold: MND is only present if a certain number of deviant neurological signs are present. This means that the severity of subfertility is not associated with milder degrees of neurological non-optimality, but only has an impact on the risk for the clinically relevant form of MND.

In line with the result of the current thesis, some studies demonstrated that subfertility-related aspects rather than ART treatment aspect may contribute to a worse child neurodevelopmental outcome.65,97,98 Zhu et al. found a delay in achieving certain motor milestones in 18 month year-old ICSI children and a modest increased risk for developmental coordination disorder (DCD) in 7-year-old children born to subfertile couples.97,98 However, the scarce literature available is less supportive towards the finding that the severity of subfertility may be the decisive factor rather than the presence of a history of subfertility per se. For instance, Zhu et al. were not able to demonstrate an effect of prolonged subfertility on the prevalence of behavioural problems in 7 to 21-year-old offspring.213 Carson et al. could not demonstrate an adverse effect of subfertility on children’s cognitive development at age 3 or 5, nor did they find evidence that a prolonged period of subfertility was associated with adverse child behaviour at age 5 and 7.116 Children born after ART had a higher mean total difficulties score than their NC peers, yet it seems that their scores mostly remained within the normal range.124 The authors speculate that the reported behavioural problems by the ART parents may be an underestimation of real problems as generally, ART parents tend to cast their children in an overly positive light and report greater warmth and parental involvement.124

A history subfertility, severe subfertility and child neurodevelopmental outcome From the results of the studies in this thesis, the question arises why subfertility-related aspects, especially the severity of subfertility, may affect the child’s neurodevelopmental outcome. Several explanations for the finding will pass in review.

Subfertility and child neurodevelopmental outcome: a socio-economical point of view Subfertile couples who may or may not seek for ART or already undergo ART are, on average, older, higher educated and more often nulliparous compared to the general obstetric population. Consequently, these patients carry additional age-related and parity-related risks or some unknown factor intrinsic risk factors potentially affecting the offspring, in an indirect or direct way, that cannot automatically be attributed to the ART treatment itself, as was demonstrated in Chapter 6. On the other hand, children born to

8

124

subfertile couples, with or without the application of ART, may generally benefit from an advantageous socioeconomic status. The relative favourable socioeconomic position (~ 40% highly educated subfertile couples) that was observed in this thesis as well may be an explanation for the number of relatively high IQ scores of the studied children.

Subfertility and child neurodevelopmental outcome: subfertility causes and obstetrical events Subfertility can be caused by a disturbance in any of several reproductive processes. It is likely that subfertile women who conceived without ART may have different fertility problems than women with fertility issues who conceive with ART. Additionally, it is expectable that each subfertility cause has its own potential consequences for child outcome, based on its origin. It is known that subfertility is associated with obstetrical problems such as pregnancy-induced hypertension or preeclampsia, placenta praevia, and increased perinatal adversities such as preterm birth, low birthweight and even perinatal deaths.59-62 The occurrence of perinatal deaths and prematurity indicate that subfertile women often have a poor overall reproductive function, as they had difficulties with both to conceive and to carry the pregnancy to term. On the other hand, one should realize that women with a history of subfertility are perhaps more closely monitored than others when finally a pregnancy is achieved, with or without the application of ART. Consequently, they have a higher pick-up rate of any potential problems, leading to an increased diagnosis of obstetric complications and the experience of stress, prior to and during the pregnancy.

In the present thesis, a relation between child developmental outcome and the underlying subfertility cause, in terms of tuba pathology, endometriosis, cervical factor or hormonal cause, male subfertility or unknown cause of subfertility could not be demonstrated. The performed subgroup analyses, however, reduced the sample size of our groups and may have affected the precision of the results. It may also be possible that other conditions than measured underlie the subfertility of the studied couples and interfered with child neurodevelopment.

Subfertility and adverse child neurodevelopmental outcome: parental and prenatal stress Another interesting thought in the explanation for the interaction of subfertility with child development is its relation to parental and perinatal stress. During fetal life, the nervous system rapidly develops, making the fetal nervous system more vulnerable to environmental influences. As discussed in Chapter 5, the data may imply that suffering from prolonged subfertility triggers the neurodevelopmental vulnerability for both sexes, and with that, nullifies the sex-specific neurodevelopmental vulnerability of the boys. An explanation for the adverse effect of TTP, which was particularly seen in girls, may be that it is due to maternal stress to which the fetus is prenatally exposed. It is likely that couples experience subfertility as a psychological burden during the period of unwanted childlessness as well during the finally achieved pregnancy. However, couples that eventually achieve pregnancy after a long period of subfertility were able to deal with this long period of unwilling childlessness and many ART treatments, are presumably more

8

General discussion

125

capable of coping with stress. Within the Groningen ART cohort study, it was previously demonstrated that a longer TTP was associated with less trait anxiety of 1-year-old children.201,242 Yet, the latter does not preclude that the offspring is affected by stress in utero as subfertile couples may underestimate the incurred stress as the joyfulness of having a child at last may overshadow the rough period of prolonged subfertility.202,203 Evidence is accumulating that maternal stress during pregnancy is associated with an increased risk of disturbance in offspring neurodevelopment.204,205 Cortisol and testosterone have been proposed as mediating hormones between maternal mental status and fetal development, as cortisol and testosterone are both the end products of two hormonal axes, and perinatal exposure to high levels of cortisol and testosterone is associated with perinatal adversities and health problems and may influence the stress response of the child itself. The altered neurochemistry due to exposure to prenatal stress may result in structural changes of the developing young brain. It is well-known that gonadal steroids play a crucial role in the sexual differentiation of the brain and its development. Animal studies have shown that prenatal stress masculinizes females and feminizes males, under the influence of androgens, impairing the reproductive capacity and reducing the sex-specific behaviour of the organism.206-210 Similar associations were seen in humans by Barrett et al. who found that exposure to prenatal life events stress was significantly associated with a longer, more masculinized anogenital distance (AGD, an indicator of prenatal androgen exposure) and masculinized play behaviour in girls, as well as a trend towards shorter, less masculine AGD in male infants but not with feminized play behaviour in boys.211,212 The authors suggest that prenatal stress may act as a non-chemical endocrine disruptor, interfering with sex-typical reproductive and neurodevelopment.211,212 The results of these studies support the hypothesis that the hormonal perturbations due to the parental subfertility-related stress mediate the neurological findings of the follow-up of the Groningen ART cohort children at 4 years: the hormonal changes may masculinize the female brain and induce more boy-like neurological vulnerability in girls, especially in the case of a long TTP, were parental stress is prolonged.

Clinical implications for subfertile couples and their children Counselling: being subfertile or suffering from subfertility? Making the transition to daily practice, what are the clinical implications of the results of this thesis? How to counsel couples with an unfulfilled child wish who may have a request for ART? Subfertile couples should be generally acknowledged as a potentially high risk group when providing prenatal care. Ideally, but maybe not the most realistic, (severe) subfertility should be prevented, which can be achieved to some extent by encouraging couples not to postpone having children, at least not until their mid-thirties. At any antenatal booking, a history of subfertility should be sought out. Counselling before any fertility treatment should include information on the potential consequences for the mother and the developing child. It should be mentioned that some of these potential consequences are inherent to the obstetric adversities itself, whereas others may be more directly related to

8

126

ART or the underlying history of subfertility. The decisions about the place of delivery for subfertile women may be influenced by this information.

Likewise, professionals must ensure that they do not unnecessarily unease the future parents, as the absolute overall risk of obstetric and neonatal complications is relatively small. Moreover, it is a challenge for the clinician to properly estimate today’s effects of ART and subfertility as ART treatment protocols have changed over the years and continue to change.

The impact of minor neurological signs The current literature available, including the current thesis, generally suggest that if ART or subfertility are related to neurodevelopmental outcome, the effect is expressed in the more subtle parameters of child neurodevelopment. Owing to their relative high incidence in a population, minor neurological signs, such as neurological non-optimality and MND, may serve as indicators for suboptimal development. In fact, the more subtle expression of the neurological condition may serve as a paradigm to understand the underlying pathophysiology in brain development related to internal or external factors. A more clinical implication of minor neurological signs is that of the presence of MND. The presence of MND, especially complex MND, is an indication for the child’s vulnerability for developing co-occurring problems, such as specific learning disorders or attention problems. It has been demonstrated that MND is associated with impaired motor performance at school age.156,158,249 MND is also associated with learning disabilities, including spelling, reading and writing impairments and impaired arithmetic skills and in a lesser extent to behavioural problems, such as attention problems, social problems and internalizing and externalizing behaviour.90,157,199,200,218,250-252 The findings of this thesis also suggests the presence of interrelationships between MND and different types of developmental disorders. A clear association was found between increased TTP and both the occurrence of complex MND and lower IQ scores at 4 years and between the presence of subfertility and behavioural problems. Additionally, cognition and behaviour were inter-related. The complex variant of MND is of clinical relevance due to its clear association with learning and behavioural problems.90,157 Moreover, from an aetiological perspective, complex MND can be considered as a borderline variant of cerebral palsy (CP), given a corresponding association between complex MND, CP and perinatal adversities.90,253 Simple MND can be considered as a mild variant of MND, reflecting a typical yet non-optimal development of the brain. Simple MND is frequently found in children and may be regarded as a minor neurological difference rather than a dysfunction. Unlike complex MND, simple MND is not evidently related to adverse conditions during early life.90 Although simple MND is considered to have little clinical relevance, it is notable that simple MND, like complex MND, is associated with an increased risk for cognitive impairments and behavioural problems in school-aged children.199,200,252 In addition to the clinical neurological condition in terms of MND, the derived optimality parameters, the fluency score and NOS, are included in the studies of this thesis. The optimality scores quantify subtle neurological deviations in the neurological

8

General discussion

127

condition of a child whereas the classification of MND is based on the presence of coherent clusters of neurological deviant signs. Clinical neurological condition and neurological optimality are highly correlated; however, a reduction in neurological optimality does not necessarily mean the presence of a neurological deviation, as the range for optimality is narrower than that for normality.160 The quantitative nature of the optimality scores makes it a suitable instrument to evaluate subtle neurological deviations for scientific research purposes rather than for clinical purposes.

As described above, minor neurological signs often do not have clinical relevance for an individual, but, on the level of a population, they may be of significance. Currently, up to 5% of newborns in Europe10 and 1% of newborns in the United States11 are born following ART. In the Netherlands, up to 3% of children is born following ART.12 Theoretically, this means that in Europe, at least one child in every classroom has been born following ART and even more children may be the result of natural conception of subfertile couples. Wittingly or unwittingly, everyone is familiar with a former classmate, who was always chosen last during gym classes. This classmate was moderately popular on school or even the misfit of the class and turned out to be a repeater. Other children made fun of him or her because of his or her clumsiness or sometimes peculiar behaviour. You may have watched these events from a distance, stood up for this person or this person was actually you. Whatever the exact situation was, it is conceivable that this particular child had complex MND, given the clumsiness in combination with the behavioural problems and the poor school performance. It goes without saying that the events in the case have an impact on the self-esteem and psychosocial functioning of young children that may persist during puberty and thereafter. The case also illustrates that subtle neurological dysfunctions may interfere with academic achievement and activities in daily life.

PGS and child neurodevelopmental outcome This thesis concludes that PGS does not seem to affect neurodevelopmental outcome of 4-year-old singletons. In contrast, PGS does seem to affect neurodevelopmental outcome of 4-year-old twins: PGS in twins was associated with a negative effect on neuromotor condition and a positive effect on the cognitive function sequential processing. This may point to the possibility that the embryo biopsy inherent to PGS is associated with differences in brain function at a later age. The findings correspond to some extent to previous findings of the PGS study. At the age of 2 years, singletons and twins were analysed together, which means that a potential effect of PGS could not be attributed exclusively to singletons or twins. It was demonstrated that children born after IVF with PGS had, on average, an approximately two points lower NOS than controls.142 At 4 years of age a similar effect was found in PGS twins, who scored approximately two points lower on the fluency score and approximately four points lower on the NOS compared to the controls. However, such an effect was not present in singletons. A negative effect of PGS was observed for singletons by Seggers et al., who investigated the same group of 4-year-old children as part of the PGS study.243 The authors found a higher frequency of received

8

128

paramedical care (speech, physical or occupational therapy) in both PGS singleton and twins compared to controls.

Within the current literature available, the PGS study is the only research project in which the effect of PGS on child outcome is studied. Other studies in the field did not differentiate between PGD and PGS, while the indication to perform one or the other is rather different. Couples who are considered for PGD treatment are at a high risk of having a child with a genetic disorder, but do not have fertility problems, whereas PGS couples are not at risk for a genetic disorder, but do have fertility problems. Hence, it is challenging to compare the findings of the PGS study to other studies. With regard to singleton children, two other groups reported on developmental outcome of singletons born after embryo biopsy.229,232,233 Nekkebroeck et al. reported on child mental, motor, socio-emotional and language development related to PGS/PGD at the age of 2. The authors demonstrated similar outcome in singletons born following PGD/PGS and singletons born following ICSI or natural conception.229,232,233 Banjeree et al. reported on health and developmental outcome, measured with the Griffiths scale, in singleton children born following PGD/PGS up to the age of 4 years. Health and developmental outcome of the PGD/PGS children was similar to that of naturally conceived peers, with the exception of locomotor development, where PGD/PGS children scored significantly lower than the controls.229,232,233

The current PGS study reported that PGS affected neurodevelopmental outcome of twins. This, however, does not match other studies.228,229 Nekkebroeck et al. were not able to demonstrate differences in mental, motor, socio-emotional and language development between 2-year-old PGS/PGD twins and twins born following ICSI or following natural conceptions The authors demonstrated similar outcome in singletons born following PGD/PGS and singletons born following ICSI or natural conception.228,229 However, the results from the current thesis match to some extent to those of the study of Liebaers et al., who found higher rates of prematurity and low birthweight in PGD/PGS multiples than in ICSI multiples and more perinatal deaths in post PGD/PGS multiple pregnancies than in post ICSI multiple pregnancies, whereas outcomes were similar for PGD/PGS and ICSI singletons.231 Yet, in a larger analysis by Desmyttere et al., including the analyses reported previously by Liebaers et al., no differences in prematurity, low birthweight and perinatal deaths were found between PGD/PGS multiples in comparison with ICSI multiples.231,234 This means that the effects of PGD/PGS on child outcome are inconsistent. Also our results do not necessarily suggest an overall adverse effect of PGS on outcome of twins, as PGS was associated with a negative effect on neuromotor condition and a positive one on sequential processing, an indicator of short-term memory. Our results rather suggest that embryo biopsy inherent to PGS affects neurodevelopmental outcome of twins in a different way than that of singletons. Generally, twins are at increased risk for developmental problems.179 Twins may be more vulnerable to external factors that may influence brain development. Moreover, they may be more subjected to such adverse circumstances. For instance, twins share the uterus, which may not be the ideal intra-uterine environment for embryonic and fetal development.

8

General discussion

129

Several explanations may be offered for a potential interference of PGS with the young, developing human brain. First, the removal of one or two blastomeres may negate any benefits that otherwise might derive from screening as it may comprise the developmental potential of the embryo with potential long-lasting consequences for the developing human. Second, the embryo biopsy inherent to PGS may not be a proper instrument to select normal embryos with a good developmental potential. One of the limitations of the technique is the limited number of chromosomes that can be analysed with FISH. Consequently, an embryo that was initially labelled as normal could be aneuploid for one or more chromosomes that were not tested. Another issue is that in assisted conception, chromosomal abnormalities in human embryos seem to be the rule rather than the exception.21,254 even in morphologically good embryos and irrespective of maternal age.255-257 This phenomenon is most likely due to mosaicism – i.e. the occurrence of more than one cell line with different genotypes in one individual.258-260 As a consequence, there is a reasonable chance that the chromosomal constitution revealed by analysis of the blastomere(s) may not be representative for the entire embryo. In fact, PGS may even favour the selection of an abnormal embryo that screens normal, over the morphologically superior embryo that screens abnormal. One factor underlying the high rate of aneuploidy and mosaicism may be the ovarian hyperstimulation. In a trial among low-risk couples that compared mild ovarian stimulation and conventional ovarian hyperstimulation, the percentage of abnormal embryos relative to the number of embryos diagnosed was 45% following mild stimulation (40 patients) compared to 63% following conventional stimulation (40 versus 33 patients; P = 0.02). In order to gain insight into chromosomal mosaicism, the authors then analysed the group of embryos in which two cells were available for diagnosis. Overall abnormality rates (abnormal and mosaic embryos) were 55% following mild and 73% following conventional ovarian stimulation (38 versus 30 patients; P = 0.046), confirming the difference in abnormality rates observed after single-cell diagnosis. Interestingly, the proportion of mosaic embryos per patient was more significantly increased following conventional ovarian stimulation (65 versus 37%; P = 0.004).261 In the present thesis, the application of ovarian hyperstimulation in both the PGS study group and the control group may be one of the explanatory factors for the similarity in neurodevelopmental outcome of the singletons.

The results described in Chapter 7 are not directly alarming, however, it may be possible that PGS, a procedure involving embryo biopsy, has any late consequences for brain function at a later age. Even though PGS is no longer practised on routine basis, the need for careful monitoring of children born following embryo biopsy remains. Embryo biopsy is still performed in the form of PGD and new techniques are continued to develop.

8

130

Methodological considerations Scientific findings must be considered within the context of their methodological limitations. Ideally, a double-blinded randomized clinical trial would be performed on the effect of assisted conception and subfertility on child outcome. However, randomly allocate (sub)fertile couples to assisted conception or natural conception is not ethical.54,262 The second best study design to thoroughly explore the effect of assisted conception and subfertility may be in the form of a control study. Difficulties to overcome in such design are namely the extended period of time it takes to conduct a long-term prospective cohort, the number of participants lost to follow up, the choice of a control group and the definition of the evaluation criteria. The methodology of the two parallel running projects of this thesis, Groningen ART cohort study and the PGS study will be addressed in the following paragraphs.

The Groningen ART cohort study The major strength of the Groningen ART cohort study lies in its unique design. First, the composition of the study groups made it possible to study the effects of separate components of assisted conception related aspects on the child’s neurodevelopmental outcome. Children born to couples who underwent ovarian hyperstimulation IVF (COH-IVF) were compared to children born to couples who underwent IVF in a modified natural cycle (MNC-IVF), without ovarian hyperstimulation. In turn, both groups were compared to a control group, a group of children born to subfertile couples who underwent no fertility treatment but conceived naturally. Parental characteristics of the latter group more closely resembled those of children born to COH-IVF and MNC-IVF, rather than a fertile control group of the general population had. Such comparison reflects the effect of ovarian hyperstimulation and the in vitro procedure together and is confounded by parental characteristics and factors associated with subfertility. As a result of the composition of the Groningen ART cohort study groups, the effects of potential confounders and a potential overestimation of the effect of ART were minimized. Second, the prospective design of our study reduced potential selection bias based on the child’s development or health, as we invited the couples of the Groningen ART cohort in the third trimester of pregnancy. Third, the assessors and supervisor were blind to prenatal and perinatal history, including the mode of conception of the ART study groups. Moreover, the likelihood that the assessors or supervisor would guess the conception mode was minimized, because all ART cohort couples had experienced subfertility. The effort that was put in the maintenance of couples with their children in the follow-up resulted in a representative study sample up until the assessment round at age 4. Throughout the study years, the postnatal attrition rate has reached only 9.3%, with an initial prenatal enrolment of 63% to 76% of eligible children.

Standardized, age-specific and highly sensitive measurements were applied to evaluate neurodevelopmental outcome in children ranged from 4 months to 4 years of age. With the Hempel examination and the IMP not only traditional signs of neurological dysfunction such as muscle tone dysregulation and motor delayed milestone achievement are assessed, but

8

General discussion

131

the quality of motor behaviour is taken into account as well. Subtle changes in neurological development are relatively unimportant on an individual level; however, on a population level they are relevant, since ART children represent a substantial part of society. Most studies however use relatively gross measurements on neurological outcome, such as the Bayley’s or Griffiths Scale of Infant Development. When instruments are not designed to study neurodevelopmental outcome in a detailed sense, potentially subtle differences between groups may remained undetected. For instance, Bouwstra et al. demonstrated an adverse effect of neonatal trans-fatty acid status on neurological development with the use of the detailed and sensitive Hempel assessment but not with the use of the Bayley’s Scale of Infant Development.164

In all studies in this thesis, several statistical approaches were applied in order to approximate the truth as closely as possible. In Chapter 2 and 7, mixed models were used as statistical model which was helpful analysing the repeated measurements on the IMP at 3 different ages and helpful analysing the child-to-mother related neurodevelopmental measurements in the PGS study. In Chapter 3, 4 and 5 univariable and multivariable regression analyses were applied to estimate the relationship between ART-related factors and subfertility aspects and neurodevelopmental outcome. In the multivariable analyses adjustments were made for potential confounders on the pathway from assisted conception, subfertility and neurodevelopmental outcome, such as birthweight, gestational age, sex, vanishing twins and parental age and educational level. The consideration to adjust for these potential confounders was twofold: a priori, based on the available literature and based on significant differences in background variables between groups. In Chapter 6 causal inference search algorithms and structural equation modelling were applied as statistical tools. Unlike traditional statistics these methods are able to unravel underlying causal mechanisms and distinguish between confounders and intermediate effects.

No study goes without criticism, therefore some points need to be discussed regarding the Groningen ART cohort study. One limitation of the study is related to its power. The original power calculation was based on the neurological outcome at 18 months.140 Moreover, the calculation was based on the neurological outcome in the ART study groups and not based on possible associations between TTP and neurological outcome. Lastly, the neurological outcome parameters to calculate the study size were based on the Hempel examination and not based on the Infant Motor Profile (IMP). At baseline, the size of the MNC-IVF group was somewhat smaller than the 64 needed for an adequate power of the study. Partly, this was compensated by larger groups of COH-IVF and Sub-NC children and fortunately, the sample size of the study groups did hardly decrease throughout the study years. In order to deal with the variety in outcome parameters and the association between TTP and neurological outcome two post hoc power analyses were performed. These analyses demonstrated that the sample size of the ART groups were slightly too small to detect ART-related effects on the IMP variation score. With respect to the TTP-effects on the prevalence of complex MND at 4 years, the sample size of the subgroup of girls was

8

132

large enough to detect relevant effects, whereas the size of the total group was somewhat underpowered.

Another limitation is that the mediation used in MNC-IVF treatments, although minimal and no hyperstimulation medication, may have caused an overestimation of the effect of the in vitro procedure or an underestimation of the effect of ovarian hyperstimulation. However, this minor confounding effect of MNC-IVF did not complicated the interpretation of the study results, as no neurodevelopmental differences between the three ART study groups were found. Another aspect of the MNC-IVF group should be mentioned. Women who underwent MNC-IVF fulfilled specific selection criteria with regards to age, BMI, previous ART treatment and the presence of an ovulatory menstrual cycle. Although we have put effort in reducing potential selection bias, it is conceivable that subtle group characteristics other than the aforementioned could have had interfere with the actual effects of assisted conception and subfertility.

Another limitation is the composition of the fertile reference group at the 2 and 4 years follow-up round. Recruitment of the reference group does not match the way in which we recruited the children of the Groningen ART cohort. Parents volunteered for participation in the reference group, which could have introduced selection bias. Parents with concerns about the health and development of their child are presumably more willing to participate in a developmental study. Moreover, the assessors and supervisor were not blinded to the group status of the reference children since this group was newly recruited.

It may be considered as a limitation that no distinction was made between IVF with ICSI and IVF without ICSI. The primary aim of the Groningen ART cohort study was to unravel the effects of ovarian hyperstimulation and the in vitro procedure on neurodevelopmental outcome and not the effect of ICSI in addition to IVF. Subdividing the ART study groups in ICSI-offspring and non ICSI-offspring would have reduced the sample size and subsequently, decreases the likelihood of finding certain effects. Once, however, attention was paid to the effect of ICSI, in an analysis in which the COH and MNC groups were pooled. This analysis did not reveal an effect of ICSI on the optimality scores and the prevalence of complex MND in 4-year-old singletons (Chapter 5).

It may also be considered as a limitation that only singletons were studied. Consequently, the results cannot be generalized to multiples, as a substantial part of those are born following controlled ovarian hyperstimulation IVF. Multiple gestation is associated with perinatal adversities and multiples are more prone to develop developmental disabilities.179 It is therefore legitimate to study developmental outcome for singletons and twins separately.

The PGS study The PGS study is the first follow-up study of a randomized controlled trial on PGS in which children were followed as long as 4 years. The random assignment of parents resulted in a strong resemblance of the two groups with respect to most background variables so that comparisons between both groups primarily reflect the effect of PGS. Additional strengths

8

General discussion

133

with regards to study design are the blinding of the assessors and supervisor to the mode of conception and the prospective design of the study. Another strength of the present study is the examination of both singletons and twins, which contributes to the generalizability of the study. Like the Groningen ART cohort study, standardized, sensitive and age-specific measurements were applied to evaluate neurodevelopmental outcome.

In both PGS and control groups children born after IVF with and without ICSI were included. In contrast to PGD where only IVF with ICSI is applied, IVF with or without ICSI are both applied whenever PGS is indicated or not. However, it could be argued that ICSI substantially differs from IVF alone. The minimization for reproductive technique (IVF or ICSI) during enrolment and the application of a sensitivity analysis dealt with the problem, indicating that the presence or absence of ICSI did not modify a potential effect of PGS.

The major limitation of the PGS study is the relatively small sample size. Power calculation was based on the number of women needed to detect a certain increase in the cumulative ongoing pregnancy rate.22 The unforeseen effect of lower pregnancy rates after PGS further reduced the number of children available for follow-up. A post hoc sample size calculation indicated that the study groups were adequately powered for the majority of outcome parameters.

Another limitation is the selective drop-out in both groups (drop-outs PGS group: higher gestational age; control group: lower educated parents). The selective loss of lower educated parents in the control group may not adequately explain the different neurodevelopmental outcome of PGS twins compared to control twins, as the selective loss was present in twins and singletons, where the selective loss in the singletons was not associated with developmental differences between PGS children and controls. Moreover, an explorative analysis revealed among others no statistically significant effects of parental educational level or gestational age.

It should be noted that participating children of both groups turned out to have a relatively high IQ. An explanation for this finding may be the relatively large proportion of highly educated parents participating in our study, as it is well known that cognitive abilities of children are positively associated with their parents education level.237 This means that our findings have to be interpreted with caution.

Future perspectives The general aim of this thesis was to evaluate the effects of assisted conception and subfertility-related aspects on neurodevelopmental outcome in preschool-aged children. The body of evidence on long-term health and development of children born to subfertile couples, with or without ART, derived from good methodological quality has to grow further. The results described in this thesis need confirmation and neurodevelopmental follow-up has to be extended to school age and the phase of adolescence to elucidate the long-term effects of ART and subfertility on the developing brain against the light of puberty-onset. A study that may confirm the findings of the current thesis, is the INeS study.263 Some years ago, this randomized clinical trial was conducted to investigate

8

134

whether COH-IVF and MNC-IVF can reduce the number of multiple pregnancy rates, but uphold effectivity, in terms of a healthy singleton, costs and patient preferences in comparison to intra-uterine insemination (IUI) with ovarian hyperstimulation (COH-IUI). Couples with unexplained or mild male subfertility were allocated to either 6 consecutive cycles of COH-IUI or 6 cycles of MNC-IVF or 3 cycles of COH-IVF with single embryo transfer (eSET) plus subsequent cryo-cycles.263 Although beyond the initial aim of the INeS study, the composition of the study groups and the randomized character of the study offers the possibility to thoroughly study the effects of ovarian hyperstimulation, the in vitro laboratory procedures and the combination of both on obstetric and child health and developmental outcome, with minimization of selection bias. The potential effect of ovarian hyperstimulation can be further explored by comparing embryonic and endometrial development after fresh and frozen-thawed embryo transfer.

Next to neurodevelopmental outcome, health outcome parameters such as cardiometabolic outcome, respiratory health and cancer deserve further attention, as recent evidence has emerged that ART children have a distinct metabolic profile that may predispose them to cardiovascular disease later in life.67,264-267

Upcoming reproductive medicine: single embryo transfer and frozen embryo replacement An important step ahead in reproductive medicine is to increase the application of eSET and with that decreasing the risk of achieving a multiple pregnancy and its associated adverse effect for the mother and the child. In eSET, a single high-quality embryo at either the cleavage or blastocyst stage of embryo development is selected from a larger number of available (high quality) embryos for transfer to the uterus.268 Currently, the number of multiple embryo transfers and multiple delivery rates are declining, while fortunately, the reduction in the number of transferred embryos has not lowered the overall pregnancy rates.10 At present, approximately 25% of embryo transfers in Europe result from eSET, whereas approximately 50% of transfers result from double embryo transfer (DET) and in less than 20% of cases, 3 or more embryos are simultaneously transferred to the uterus.10 The current trend of transferring fewer embryos has resulted in more embryos being available for freezing (frozen embryo replacement, FER). The increased use of FER has intensified the awareness of the safety aspects of this procedure and also offers the opportunity to gain more insight in the effect of the supraphysiologic hormonal milieu resulting from ovarian stimulation. In contrast to fresh embryos, cryopreservation embryos are not exposed to a hormonally loaded endometrium, presumably resulting in more optimal endometrial circumstances. Support for this hypothesis is promoted by Wennerholm et al., who systematically reviewed the literature available on child outcome after cleavage stage embryo cryopreservation up until 2008.175 The authors concluded that children born after FER showed similar or even improved perinatal outcomes compared to children born following fresh cycles. Moreover, the proportion of low birthweight and prematurity after FER is similar to that of singletons born after from spontaneous conception. After the review of Wennerholm et al., two Scandinavian large cohort studies

8

General discussion

135

were performed on perinatal outcome of children born following FER in comparison to peers born following fresh embryo transfer. Like Wennerholm et al., the studies concluded that FER does not adversely affect perinatal outcome and the outcomes are similar or even better, particularly regarding fetal growth, compared to those of children born following fresh embryo transfer.176,269 However, there are concerns about an increased risk of being large for gestational age (LGA) after FER in comparison to both fresh embryo transfers and spontaneous conception.269-271 It is hypothesized that the cryopreservation techniques may induce changes in the developmental processes in the early embryo stages and hence in the intra-uterine growth potential. The possible asynchrony between the endometrium and embryo that occurs in case of FER may also play a role in the occurrence of higher rates of LGA children after FER as it may alter the subsequent trajectory of fetal growth and development.33,271 The increased risk of being LGA after birth following FER deserves further attention, given the growing trend towards single embryo-transfer policy and the resulting increase in freezing of the non-transferred embryos.

The future of preimplantation genetic screening In modern society, science is driven at top speed and also in reproductive medicine new developments in assisted conception take place. Such fertility treatment changes may have an impact on the various outcomes and complicate the interpretation of study findings. It appears that over time, the overall outcomes for children born following ART has improved.272 The improvement may be the result of the enhancement of clinical and laboratory skills and the increasing use of eSET or the current accessibility of ART for subfertile couples. With the introduction of new treatment regimes, responsibility follows to carefully monitor and warrant its safety. Ideally, safety evaluation is performed prospectively. However, from an ethical point of view, such strategy is usually not possible in reproductive medicine. Consequently, theoretically well-thought techniques may turn out differently in daily practice, as in the case of PGS.22,23 PGS as applied in the PGS study in this thesis is currently no longer practiced, however, embryo biopsy is still applied in the form of PGD. Moreover, the techniques used in PGS and PGD are subjected to developments to improve its accuracy in terms of the timing of the biopsy (i.e. cleavage or blastocyst stage), the material used for screening (embryonic cells, the polar body or trophectoderm tissue) and the number of chromosomes that are needed to be screened. At present, it is possible to screen for aneuploidies using all 23 pairs of chromosomes of a single cell using more advanced genetic analysis such as comparative genomic hybridization arrays or single nucleotide polymorphism arrays.273 Future research should focus on the effectiveness of these new techniques. And potential adverse effects on health and development of the offspring has to be identified before large scale implementation of new techniques.

8

136

Extending long-term follow-up: are subfertile couples becoming grandparents? It is important to investigate whether adverse health and developmental outcome of children manifest later in life. It is to be expected that in the coming years, the follow-up period of children born to subfertile couples who conceived spontaneously or with the help of ART will be expanded to adolescence and adulthood as an increasing number of this next generation has reached the fertile age. The first person born following ART, Louise Brown, is now 36 years old. She had her first child at the age of 27 years, a son named Cameron, who was conceived naturally within 6 months TTP.274,275 Whether or to what extent some subfertility problems are heritable may depend on the type of subfertility. For instance, Louise Brown’s mother underwent IVF because of blocked fallopian tubes, which is usually an acquired condition and may therefore not influence the fertility patterns of the offspring.8,9 However, other fertility outcomes may be expected in case of unknown subfertility, male subfertility or subfertility due to polycystic ovarian syndrome or endometriosis, disorders that may already be inheritable by itself. Boys conceived using ICSI inherit the Y-chromosome of their fathers, which may contain abnormalities that may be related to the underlying male subfertility of the father. Whether ART interferes with the fertility potential of the offspring is an interesting question that needs further research. In theory, ART may interfere with the early development of the endocrine organs due to hormonal stimulation, which may be reflected by disturbances in the pubertal development. From the scarce literature available exploring this topic so far, it can be deduced that ART does not appear to affect the onset of puberty.75,85,86 However, there are some suggestions that ART may interfere at a different level in puberty, such as delayed breast development,87 advanced bone age and higher serum levels of luteinizing hormone (LH) and dehydroepiandrosterone sulphate (DHEAS, a steroid hormone, that functions as a metabolic intermediate in the biosynthesis of the androgen and oestrogen)85 in girls and lower serum testosterone concentrations and higher LH/testosterone ratio in boys.88 In the latter study, the deviant hormonal levels were only seen in boys born following ICSI, but not in boys born following IVF or natural conception, suggesting that the ICSI boys may have inherited the subtle impairment of Leydig cells from their fathers.88 However, another study reported normal male prepubertal development after ICSI, in terms of penile and testicular growth as well as Sertoli cell function.276 During puberty, the same group of ICSI boys showed testosterone levels comparable to NC boys.277,278 Moreover, ICSI boys from fathers with severely compromised spermatogenesis showed testosterone levels comparable to those from fathers with normal spermatogenesis.277,278

8

General discussion

137

Concluding remarks The following conclusions can be drawn from the studies described in this thesis: • Ovarian hyperstimulation, the in vitro laboratory procedures, or a combination of

both are not associated with worse neurodevelopmental outcome - in terms of movement variation, neuromotor development, cognitive and behaviour development - in singleton children aged 4 months to 4 years.

• The severity of subfertility, reflected by an increased TTP, rather than the presence or specific underlying causes of subfertility are associated with minor neurological dysfunction at 2 and 4 years.

• The presence and especially the severity of subfertility affected cognitive and behavioural development at 4 years, with maternal age at child conception and maternal educational level as the most important confounders.

• Neurodevelopmental outcome – in terms of neuromotor development, cognitive and behaviour development – in singletons born following IVF with PGS is similar to that of singletons born after IVF without PGS at age 4. However, PGS seems to be associated with altered neurodevelopment in twins.

The main conclusion that subfertility-related aspects rather than ART-related aspects contribute to a worse neurodevelopmental outcome of preschool-aged children, with the severity of subfertility being the decisive factor, underlines the importance of long-term follow-up of health and development of children born to subfertile couples, without and with the application of ART. In a society where maternal age at child birth, subfertility and the application of ART are steadily increasing, it is of utmost importance to monitor and to improve the safety and health of subfertile couples in their quest for parenthood and of their future offspring.

8

138

Summary

139

CHAPTER 9 Summary

140

Five percent of newborns in Europe are born following assisted reproductive techniques (ART), a steadily increasing number. Consequently, development and health of children born following ART is of general significance. ART is associated with perinatal adversities such as low birthweight and preterm delivery. Nevertheless, ART has not been associated with adverse neurodevelopmental outcome during the first postnatal years. However, this does not preclude an effect of ART on long-term neurological development. Neurodevelopmental disorders may first emerge later in life, as a result of the continuing structural and functional development of the brain during childhood.

In this thesis the neurodevelopmental outcome of children born following ART is evaluated up until 4 years of age by exploring the influence of specific factors involved in assisted conception, such as ovarian hyperstimulation, the in vitro laboratory procedures and subfertility-related aspects. The subfertility aspects are threefold: the presence of a history of subfertility, the underlying cause of subfertility and the duration of subfertility in terms of time to pregnancy (TTP), as a proxy for the severity of subfertility.

Chapter 1 gives a general and concise overview of the literature available on child health and developmental outcome after ART. The overview indicates that ART does not seem to be associated with adverse neurodevelopmental outcome during the first postnatal years. However, insufficient information is available on the long-term consequences of assisted conception and subfertility-related aspects on a child’s neurodevelopment. Chapter 1 also describes the concept of brain development and vulnerability and introduces the applied fertility treatment techniques in the studies of the thesis. Finally the chapter introduces the two projects, the Groningen ART cohort study and the PGS (preimplantation genetic screening) study, on which the studies of the thesis are based.

Part I: The Groningen ART cohort study Part I evaluates primarily the potential effects of ovarian hyperstimulation, the in vitro laboratory procedures, or a combination of both on neurodevelopmental outcome in preschool-aged children. Secondly, part I evaluates the potential effects of subfertility on neurodevelopmental outcome.

In Chapter 2 the potential effect of ART on neurodevelopmental outcome, expressed in terms of movement variation, was evaluated in the children of the Groningen ART cohort at the ages of 4, 10 and 18 months. All children were singletons, born following in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) with conventional controlled ovarian hyperstimulation (COH-IVF, n=68), after IVF in a modified natural cycle (MNC-IVF, n=57) or after natural conception born to subfertile couples (Sub-NC, n=90). Neurodevelopmental outcome was measured with the Infant Motor Profile (IMP), resulting in a total IMP score and five domain scores: variation, variability, symmetry, fluency and performance. Primary outcome was the domain score variation (i.e. movement repertoire size), a parameter reflecting the integrity of cortical connectivity. The main result of the study was that COH-IVF children had a significantly lower mean variation score than MNC-IVF children. However, a similar difference was not observed between the two IVF groups

9

Summary

141

and the Sub-NC group. The study concludes that no clear effect of ovarian hyperstimulation and the in vitro procedure on movement variation throughout infancy could be demonstrated.

Chapter 3 describes the neurological condition of the children of the Groningen ART cohort at the age of 2 years in terms of neurological optimality (neurological optimality score [NOS], fluency score) and the prevalence of minor neurological dysfunction (MND). Special attention was paid to potential effects of the presence and the underlying cause of subfertility on the outcome parameters. Singletons born following COH-IVF (n=66), MNC-IVF (n=56), Sub-NC (n=87) and after natural conception to fertile couples (reference group, n=101) were assessed with the Hempel examination. Primary outcome was the fluency score, as fluency of movements is easily affected by subtle dysfunction of the nervous system. The main results of the study were similar fluency scores, NOS and MND prevalence in the groups. However, the fluency score and NOS of the three subfertile groups were higher, and the prevalence of MND was lower than those in the reference group. The study concludes that the neurological condition of 2-year-olds born after ART is similar to that of children of subfertile couples conceived naturally. Moreover, subfertility does not seem to be associated with a worse neurological outcome.

Chapter 4 describes the effect of the severity of subfertility, in terms of TTP, on the neurological condition of the children of the Groningen ART cohort at 2 years in terms of MND, assessed with the Hempel examination. Participants were singleton children born to subfertile couples (n=209). The main result of the study was that TTP of children with MND was significantly longer than that of children without MND. After correction for gestational age, parental age and parental level of education (multivariable logistic regression analyses), the association remained statistically significant. The study concludes that increased TTP was associated with suboptimal neurological development in 2-year-olds. This suggests that subfertility and its determinants are involved in the genesis of neurodevelopmental problems.

Chapter 5 describes the effect of both ART-related aspects and the three previously mentioned subfertility aspects on the neurological condition of the children of the Groningen ART cohort at the age of 4 years, again, in terms of neurological optimality (NOS, fluency score) and the prevalence of complex MND. Specific attention was paid to the possibility of sex-specific effects of ART and subfertility. Four-year-old singletons born to subfertile parents (subfertile group, n=195), including singletons born following COH-IVF (n=63), MNC-IVF (n=53) and Sub-NC (n=79) and newly recruited 4-year-old singletons born to fertile parents after natural conception (reference group, n=98) were studied. Neurological development was evaluated with the Hempel examination, with the fluency score being the primary outcome parameter. The main results of the study were that the fluency score, NOS and the prevalence of complex MND were similar in COH-IVF, MNC-IVF and Sub-NC children. Neurological condition of children born to subfertile parents was similar to that of children of fertile parents and was independent of the underlying cause of subfertility. No statistically significant associations were found between TTP and the fluency

9

142

score and NOS. However, a positive correlation was found between TTP and the prevalence of complex MND; a correlation which could be attributed to girls, in whom an evident positive correlation was present. A similar association was absent in boys. The study concludes that ovarian hyperstimulation, the in vitro procedure and the combination of both, a history of subfertility per se and the underlying cause of subfertility were not associated with worse neurological outcome in 4-year-old singletons. However, increased TTP was associated with an increased risk for the complex form of MND, especially in girls. This suggests that rather the severity of subfertility than its presence or IVF-components affects neurological outcome. Moreover, girls may be neurologically more vulnerable for the effect of severity of subfertility than boys.

Chapter 6 describes an explorative approach on the effect of both the ART-related aspects and the three subfertility aspects on the cognitive and behavioural development of the children of the Groningen ART cohort at the age of 4 years. Four-year-old singletons born to subfertile parents (subfertile group, n=195), including singletons born following COH-IVF (n=63), MNC-IVF (n=53) and Sub-NC (n=79) and newly recruited 4-year-old singletons born to fertile parents after natural conception (reference group, n=98) were studied. The primary cognitive outcome parameter was a total intelligence quotient (IQ), measured with the Kaufman Assessment Battery for Children, second edition (K-ABC-II); the behavioural outcome parameter was the total problem T-score, measured with the Child Behavior Checklist (CBCL). Causal inference search algorithms and structural equation modelling were applied as statistical tools. Unlike traditional statistics these methods are able to unravel underlying causal mechanisms and distinguish between confounders and intermediate effects. The main results of the study were the absence of direct or indirect causal effects of ovarian hyperstimulation or the in vitro procedure on cognitive and behavioural outcome in 4-year-old singletons born to subfertile couples. However, direct negative causal effects were found of the severity of subfertility in terms of time to pregnancy (TTP) on cognition and of the presence of subfertility on behaviour. Both effects were confounded by maternal age at child conception and maternal educational level. Additionally, cognition and behaviour were directly related to one another. The study concludes that suffering from subfertility per se, and especially from more severe subfertility – which by itself is associated with higher age and high educational level of the mother – negatively affects the child’s cognitive and behavioural outcome.

Part II: The Groningen PGS study Part II evaluates the effect of preimplantation genetic screening (PGS) on neurodevelopmental outcome, in terms of neuromotor, cognitive and behavioural development of 4-year-old children.

Chapter 7 describes the results of this prospective, assessor-blinded follow-up study on children born to women who participated in a multicentre randomized clinical trial (RCT) on the effect of IVF with or without PGS. Forty-nine children (31 singletons, 9 sets of twins) were born following IVF with PGS and 64 children (42 singletons, 11 sets of twins) were born

9

Summary

143

following IVF without PGS (controls). Assessments were the Hempel examination, resulting in a fluency score (primary outcome parameter), NOS and the prevalence of complex MND; the K-ABC-II, resulting in a total IQ score and the CBCL, resulting in a total problem T-score. The main result of the study was that neurodevelopmental outcome of PGS children was similar to that of controls. However, PGS affected neurodevelopmental outcome of twins in a different way than that of singletons. The fluency score and the NOS of PGS twins were significantly lower than those of control twins. In addition, PGS in twins was associated with a higher sequential IQ score. The study concludes that PGS does not seem to affect neurodevelopmental outcome of 4-year-old singletons, however, our data suggest that it may be associated with altered neurodevelopment in twins.

Chapter 8 discusses the findings of the studies in this thesis. The thesis concludes that ART-related aspects such as ovarian hyperstimulation, the in vitro laboratory procedures, or a combination of both were not associated with worse neurodevelopmental outcome – in terms of movement variation, neuromotor function, cognition and behaviour – in singleton children aged 4 months to 4 years. Additionally, a history of subfertility per se and the underlying cause of subfertility were not associated with worse neurological outcome in 2 and 4-year-old singleton children. However, increased TTP was associated with an increased risk for MND at 2 years and an increased risk for complex MND at 4 years. Moreover, subfertility per se, and especially more severe subfertility – which by itself is associated with higher age and high educational level of the mother – negatively affected the child’s cognitive and behavioural outcome. Furthermore, neurodevelopmental outcome in singletons born following IVF with PGS was similar to that of singletons born after IVF without PGS at age 4. PGS in twins was associated with a negative effect on neuromotor condition and a positive effect on sequential processing IQ. This may point to the possibility that the embryo biopsy inherent to PGS is associated with differences in brain function at a later age.

The major strengths of the Groningen ART cohort study is its prospective design, the unique composition of the study groups and the use of standardized, age-specific and sensitive tools used to evaluate neurodevelopmental outcome. Additional strengths are the blinding of our assessors to the mode of conception of the Groningen ART cohort children and the less than 10% post-natal attrition rate throughout the years. A limitation of the Groningen ART cohort study is the composition of the fertile reference group which does not match the way in which we recruited the children of the Groningen ART cohort. Moreover, the results cannot be generalized to multiples, as we studied singletons only.

The major strengths of the PGS study are the random assignment of the participants, the blinding of the assessors to the mode of conception and the prospective design of the study. The PGS study is the first follow-up study of a RCT on PGS in which children were followed as long as 4 years. Limitations are the relative small size of our study groups and the selective drop-out in both groups.

9

144

Future research on long-term health and development of children born to subfertile couples, with or without ART, has to continue. Health and neurodevelopmental follow-up has to be extended to school-aged children and adolescents to elucidate the long-term effects of ART and subfertility on the developing brain and body against the light of puberty-onset. Subsequently, it is also interesting to investigate the association between ART and parental subfertility and the offspring’s fertility potential. Attention should also be paid to the effectiveness and safety of new techniques in assisted conception. At present, new techniques are being developed to improve the accuracy of PGS in terms of the timing of the biopsy (i.e. cleavage or blastocyst stage), the material used for screening (embryonic cells, the polar body or trophectoderm tissue) and the number of chromosomes that are needed to be screened. Another upcoming technique is frozen embryo replacement (FER). As single embryo transfer (eSET) is currently increasingly often applied, more embryos are available for freezing and a growing number of children are already born following FER. Health and development of these children needs to be monitored, as concerns have been raised about an increased risk of being large for gestational age (LGA) after FER in comparison to both fresh embryo transfers and spontaneous conception.

The results of this thesis underline the importance of long-term follow-up of health and development of children born to subfertile couples, without and with the application of ART. Monitoring and improving the safety and health of subfertile couples and their children is necessary, especially in today’s society were maternal age at child birth, subfertility and the application of ART are increasing and fertility treatment techniques are continuously being developed and renewed.

9

Nederlandse samenvatting (Summary in Dutch)

145

CHAPTER 10 Nederlandse samenvatting (Summary in Dutch)

146

Vijf procent van alle pasgeborenen in Europa wordt geboren na hulp bij voortplanting (kunstmatige voortplanting, assisted reproductive techniques, ART), een percentage dat nog steeds toeneemt. Dit maakt de ontwikkeling en gezondheid van kinderen die geboren zijn na ART van algemeen maatschappelijk belang. ART is geassocieerd met slechtere perinatale uitkomsten zoals een te laag geboortegewicht en vroeggeboorte. Toch lijkt ART niet geassocieerd te zijn met een slechtere neurologische ontwikkeling gedurende de eerste postnatale jaren. Echter, dit sluit een effect van ART op de lange termijn ontwikkeling niet uit. Ontwikkelingsstoornissen kunnen later in het leven tot uiting komen, als gevolg van de continue structurele en functionele ontwikkeling van het brein in de kindertijd.

In dit proefschrift wordt de neurologische ontwikkeling tot op de leeftijd van 4 jaar geëvalueerd van kinderen die geboren zijn na ART. Dit wordt gedaan door de invloed van verschillende factoren te bestuderen die verband houden met hulp bij voortplanting, zoals ovariële hyperstimulatie, de in vitro laboratorium procedures en de potentiële effecten van subfertiliteit. Bij de subfertiliteit wordt specifiek aandacht besteed aan drie aspecten: de aanwezigheid van een voorgeschiedenis van subfertiliteit, de onderliggende oorzaak van de subfertiliteit en de duur van de subfertiliteit in termen van tijd tot zwangerschap (time to pregnancy, TTP), als maat voor de ernst van de subfertiliteit.

Hoofdstuk 1 geeft een algemeen en beknopt overzicht van de beschikbare literatuur over de gezondheid en ontwikkeling van kinderen die geboren zijn na ART. Dit overzicht geeft aan dat ART niet geassocieerd lijkt te zijn met een slechtere ontwikkelingsneurologische uitkomst gedurende de eerste levensjaren. Echter, we zijn nog niet goed op de hoogte van de consequenties van ART en subfertiliteit-gerelateerde factoren op de neurologische ontwikkelingen van het kind op de lange termijn. In hoofdstuk 1 wordt ook het concept van hersenontwikkeling en –gevoeligheid behandeld en worden de verschillende voortplantingstechnieken die in dit proefschrift aan bod komen, beschreven. Tot slot worden de twee projecten, de Groningen ART cohort studie en de PGS studie, waarop het onderzoek van dit proefschrift is gebaseerd, geïntroduceerd.

Deel I: De Groningen ART cohort studie Deel I evalueert primair de potentiele effecten van ovariële hyperstimulatie, de in vitro laboratorium procedures of de combinatie van beide op de neurologische ontwikkeling van kinderen in de voorschoolse leeftijd. Daarnaast evalueert deel I de potentiële effecten van subfertiliteit op de neurologische ontwikkeling.

In hoofdstuk 2 worden de potentiele effecten van ART op de neurologische ontwikkeling, uitgedrukt in variatie in beweging, geëvalueerd in kinderen van het Groningen ART cohort op de leeftijd van 4, 10 en 18 maanden. Alle kinderen waren eenling en geboren na in-vitrofertilisatie (IVF) of intracytoplasmatische sperma-injectie (ICSI) met conventionele gecontroleerde ovariële hyperstimulatie (COH-IVF, n=68), na IVF in de gemodificeerde eigen menstruele cyclus (MNC-IVF, n=57) of na natuurlijke conceptie uit subfertiele ouderparen (Sub-NC, n=90). De neurologische ontwikkeling werd gemeten met de Infant Motor Profile (IMP), resulterend in een IMP totaal score en vijf domeinscores: variatie, variabiliteit, 10


147

symmetrie, vloeiendheid en uitvoering. De primaire uitkomstmaat was de domeinscore variatie (d.i. de grootte van het bewegingsrepertoire), een parameter die de integriteit van de corticale connectiviteit representeert. Het belangrijkste resultaat van de studie was dat COH-IVF kinderen een significant lagere gemiddelde variatie score hadden dan MNC-IVF kinderen. Een soortgelijk verschil werd niet geobserveerd tussen de twee IVF groepen in vergelijking met de Sub-NC groep. De studie concludeert dat er geen duidelijk effect kon worden aangetoond van ovariële hyperstimulatie en de in vitro laboratorium procedures op de bewegingsvariatie van zuigelingen.

Hoofdstuk 3 beschrijft de neurologische conditie van kinderen van het Groningen ART cohort op 2-jarige leeftijd in termen van neurologische optimaliteit (neurologische optimaliteitsscore [NOS], score voor vloeiend bewegen [fluency score]) en de prevalentie van lichte neurologische disfuncties (minor neurological dysfunction, MND). Er is speciale aandacht besteed aan de potentiële effecten van de aanwezigheid van een voorgeschiedenis van subfertiliteit en de onderliggende oorzaak van subfertiliteit. Kinderen geboren uit subfertiele ouderparen (subfertiele groep, n=209) na COH-IVF (n=66), MNC-IVF (n=56), Sub-NC (n=87) en na natuurlijke conceptie uit fertiele ouderparen (referentiegroep, n=101) werden onderzocht met behulp van het neurologisch onderzoek volgens Hempel. De primaire uitkomstmaat was de fluency score, aangezien de vloeiendheid van bewegingen al is aangedaan wanneer er slechts sprake is van subtiele dysfunctie van het zenuwstelsel. Het belangrijkste resultaat van de studie was dat de fluency score, NOS en de prevalentie van MND gelijk waren in de drie subfertiele groepen. Echter, de fluency score en NOS van de drie subfertiele groepen waren hoger en de prevalentie van MND was lager in vergelijking met de referentie groep. De studie concludeert dat de neurologische conditie van 2-jarigen geboren na ART vergelijkbaar is met die van kinderen geboren na natuurlijke conceptie uit subfertiele ouderparen. Daarnaast lijkt subfertiliteit niet geassocieerd te zijn met een slechte neurologische uitkomst.

Hoofdstuk 4 beschrijft het effect van de mate van subfertiliteit, in termen van TTP, op de neurologische conditie, van de Groningen ART cohort kinderen op de leeftijd van 2 jaar. De neurologische conditie werd uitgedrukt in termen van MND en geëvalueerd met behulp van het Hempel onderzoek. Deelnemers waren eenlingen geboren uit subfertiele ouderparen (n=209). Het belangrijkste resultaat van de studie was dat TTP van ouderparen van kinderen met MND significant langer was dan dat van ouders van kinderen zonder MND. Na statistische correctie voor gestatieduur, de leeftijd en het opleidingsniveau van de ouders (multivariabele logistische regressie analyse) bleef de associatie statistisch significant. De studie concludeert dat een langere TTP geassocieerd is met een suboptimale neurologische ontwikkeling in 2-jarige kinderen. Dit suggereert dat subfertiliteit en haar determinanten mogelijk betrokken zijn in het ontstaan van neurologische ontwikkelingsproblematiek.

Hoofdstuk 5 beschrijft de effecten van aspecten van ART en die van subfertiliteit op de neurologische conditie van de kinderen van het Groningen ART cohort op de leeftijd van 4 jaar, wederom in termen van neurologische optimaliteit (NOS, fluency score) en deze keer de prevalentie van complexe MND. Er werd speciale aandacht besteed aan eventuele sekse- 10

148

specifieke effecten van ART en subfertiliteit. Vierjarige eenlingen geboren uit subfertile ouderparen (subfertiele groep, n=195), inclusief eenlingen geboren na COH-IVF (n=63), MNC-IVF (n=53) en Sub-NC (n=79) en nieuw gerekruteerde 4-jarige eenlingen die na natuurlijke conceptie geboren werden uit fertiele ouders (referentiegroep, n=98), werden onderzocht. De neurologische ontwikkeling werd wederom geëvalueerd met het Hempel onderzoek, met de fluency score als primaire uitkomstmaat. De belangrijkste resultaten van de studie waren dat de fluency score, NOS en de prevalentie van complexe MND gelijk waren in COH-IVF, MNC-IVF en Sub-NC kinderen. De neurologische conditie van kinderen geboren uit subfertiele ouders was vergelijkbaar met die van kinderen geboren uit fertiele ouders en was onafhankelijk van de onderliggende subfertiliteitsoorzaak. Er werden geen statistisch significante associaties gevonden tussen TTP en de fluency score of NOS. Er bestond echter wel een positieve correlatie tussen TTP en de prevalentie van complexe MND; een correlatie die kon worden toegeschreven aan de meisjes, bij wie de correlatie evident was. Een vergelijkbare correlatie werd niet bij de jongens gevonden. De studie concludeert dat ovariële hyperstimulatie, de in vitro procedure en de combinatie van beide, een voorgeschiedenis van subfertiliteit en de onderliggende subfertiliteitsoorzaak niet waren geassocieerd met een slechtere neurologische uitkomst op 4-jarige leeftijd. Een langere TTP was echter wel geassocieerd met een verhoogd risico op de complexe vorm van MND, in het bijzonder bij meisjes. Dit suggereert dat met name de ernst van de subfertiliteit en niet de aanwezigheid van subfertiliteit an sich of IVF-componenten de neurologische uitkomst negatief beïnvloeden. Meisjes zijn mogelijk vatbaarder voor het effect van ernstige subfertiliteit dan jongens.

Hoofdstuk 6 beschrijft een statistisch exploratieve benadering van de effecten van aspecten van ART en die van subfertiliteit op de cognitieve en gedragsmatige ontwikkeling van de kinderen van het Groningen ART cohort op de leeftijd van 4 jaar. Vierjarige eenlingen geboren uit subfertiele ouderparen (subfertiele groep, n=195), inclusief eenlingen geboren na COH-IVF (n=63), MNC-IVF (n=53) en Sub-NC (n=79) en nieuw gerekruteerde 4-jarige eenlingen die na natuurlijke conceptie geboren werden uit fertiele ouders (referentiegroep, n=98) werden bestudeerd. De primaire cognitieve uitkomstmaat was het totale intelligentie quotiënt (IQ), gemeten met de Kaufman Assessment Battery for Children, tweede editie (K-ABC-II); de uitkomstmaat voor gedrag was de totale probleem T-score, gemeten met de Child Behavior Checklist (CBCL). Exploratieve causale inferentie zoekalgoritmes en structurele vergelijkingsmodellen werden toegepast als statistische technieken. Anders dan traditionele statistiek maken deze technieken het mogelijk onderliggende causale mechanismen te ontrafelen en maken ze onderscheid in confounders en intermediërende effecten. De belangrijkste resultaten van de studie waren de afwezigheid van directe of indirecte causale effecten van ovariële hyperstimulatie en de in vitro laboratorium procedure op de cognitieve en gedragsmatige ontwikkeling in 4-jarige eenlingen geboren uit subfertiele ouders. Wel werden er directe negatieve causale effecten gevonden van zowel de ernst van de subfertiliteit in termen van TTP op de cognitie en van de aanwezigheid van een voorgeschiedenis van subfertiliteit op het gedrag. Beide effecten 10


149

werden vertekend door de leeftijd van de moeder op het moment van conceptie en het opleidingsniveau van de moeder. Daarnaast waren cognitie en gedrag direct gerelateerd aan elkaar. De studie concludeert dat de aanwezigheid van subfertiliteit en met name die van ernstige subfertiliteit – welke op zichzelf geassocieerd zijn met een hogere leeftijd en een hoger opleidingsniveau van de moeder – van negatieve invloed is op de cognitieve en gedragsmatige ontwikkeling van het kind.

Deel II: De PGS studie Deel II evalueert het effect van preïmplantatie genetische screening (PGS) op de neurologische ontwikkeling, in termen van neuromotorische, cognitieve en gedragsmatige ontwikkeling, van 4-jarige kinderen. Hoofdstuk 7 beschrijft deze prospectieve, onderzoekers-geblindeerde follow-up studie over kinderen geboren uit vrouwen die meededen in een multicenter gerandomiseerde klinische trial (RCT) naar het effect van IVF met of zonder PGS. Negenenveertig kinderen (21 eenlingen, 9 tweelingen) werden geboren na IVF met PGS en 64 kinderen (42 eenlingen, 11 tweelingen) werden geboren na IVF zonder PGS (controlegroep). De onderzoeksmethoden waren het Hempel onderzoek, resulterend in de fluency score (vloeiendheid van bewegingen, primaire uitkomstmaat), de NOS en de prevalentie van complexe MND; de K-ABC-II, resulterend in de totale IQ score en de CBCL, resulterend in de totale probleem T-score. Het belangrijkste resultaat van de studie is dat de neurologische ontwikkeling van PGS kinderen vergelijkbaar is met die van de controle kinderen. Echter, PGS had een andere invloed op tweelingen dan op eenlingen. De fluency score en de NOS van PGS tweelingen waren significant lager dan die van controle tweelingen. Daarnaast was PGS in tweelingen geassocieerd met een hogere sequentiële IQ score. De studie concludeert dat PGS niet van invloed lijkt te zijn op de neurologische ontwikkeling van 4-jarige eenlingen, maar dat deze mogelijk wel is geassocieerd in een andere neurologische ontwikkeling van tweelingen. Hoofdstuk 8 bespreekt de bevindingen van de studies in dit proefschrift. De conclusie van dit proefschrift is dat ART gerelateerde aspecten zoals ovariële hyperstimulatie, de in vitro laboratorium procedures of de combinatie van beide niet geassocieerd zijn met een slechtere ontwikkelingsneurologische uitkomst – in termen van bewegingsvariatie, neuromotorisch functioneren, cognitie en gedrag – in eenling kinderen op de leeftijd van 4 maanden tot en met 4 jaar. Daarnaast zijn een voorgeschiedenis van subfertiliteit en de onderliggende oorzaak van subfertiliteit niet geassocieerd met een slechtere neurologische uitkomst in 2- en 4-jarige eenlingen. Echter, een langere TTP is wel geassocieerd met een verhoogd risico op het ontwikkelen van MND op 2-jarige leeftijd en complexe MND op 4-jarige leeftijd. Verder is subfertiliteit, in het bijzonder meer ernstige subfertiliteit – welke op zichzelf is geassocieerd met een hogere leeftijd en een hoger opleidingsniveau van de moeder - van negatieve invloed op de cognitieve en gedragsmatige ontwikkeling van 4-jarige eenlingen. De neurologische ontwikkeling van 4-jarige eenlingen die geboren zijn na IVF met PGS is vergelijkbaar met die van 4-jarige eenlingen die geboren zijn na IVF zonder 10

150

PGS. In tweelingen is PGS van negatieve invloed op het neuromotorisch functioneren van het kind en van positieve invloed op het sequentiële verwerken door het brein. Dit duidt op de mogelijkheid dat embryobiopsie inherent aan PGS is geassocieerd met verschillen in hersenontwikkeling op oudere leeftijd.

De sterkste punten van de Groningen ART cohort studie zijn haar prospectieve design, de unieke samenstelling van de onderzoeksgroepen en het toepassen van gestandaardiseerde, leeftijdsspecifieke en gevoelige meetinstrumenten om de neurologische ontwikkeling in kaart te brengen. Andere sterke punten van de studie zijn de blindering van de onderzoekers voor de conceptiemethode binnen de ART cohort deelnemers en de geringe postnatale uitval van deelnemers door de jaren heen (minder dan 10%). Een beperking van de studie is de wijze waarop de fertiele referentiegroep is samengesteld; deze kwam niet overeen met die van de ART cohort groepen. Daarnaast zijn de resultaten niet te generaliseren naar tweelingen, aangezien alleen eenlingen zijn onderzocht.

De sterkste punten van de PGS studie zijn de randomisatie van de deelnemers, de blindering van de onderzoekers voor de conceptiemethode en het prospectieve design van de studie. De PGS studie is de eerste follow-up studie van een RCT over PGS waarin kinderen zo lang, dat wil zeggen tot op 4-jarige leeftijd, werden gevolgd. Beperkingen zijn de relatief kleine groepsgroottes en de selectieve uitval in beide groepen.

Toekomstig onderzoek naar de lange termijn gezondheid en ontwikkeling van kinderen geboren uit subfertiele ouderparen, met of zonder ART, moet worden gecontinueerd. Dergelijk onderzoek moet worden doorgetrokken naar de schoolleeftijd en adolescentie om de lange termijn effect van ART en subfertiliteit op het zich ontwikkelende brein en het lichaam onder invloed van de puberteit op te helderen. Daarnaast verdient de potentiele associatie tussen ART en ouderlijke subfertiliteit en het voortplantingspotentieel van de nakomelingen nadere aandacht. Er moet ook aandacht zijn voor de effectiviteit en veiligheid van nieuwe technieken binnen de vruchtbaarheidsbehandelingen. Momenteel worden er nieuwe PGS technieken ontwikkeld om de nauwkeurigheid van PGS te verbeteren, in termen van timing van de biopsie (klievingsdeling- of blastocyststadium), het gebruikte screeningsmateriaal (embryonale cellen, het poollichaampje of het trophectoderm) en het aantal chromosomen dat gescreend kan worden. Een andere opkomende techniek is de terugplaatsing van ingevroren-ontdooide embryo’s (frozen embryo replacement, FER). Door het toenemende terugplaatsen van slechts één embryo (single embryo transfer, eSET) zijn er meer embryo’s beschikbaar voor cryopreservatie en wordt een groeiend aantal kinderen geboren na FER. De gezondheid en ontwikkeling van deze kinderen moet goed in de gaten gehouden worden, aangezien er zorgen zijn over het verhoogde risico op een te hoog geboortegewicht voor de zwangerschapsduur (large for gestational age, LGA) in vergelijking met kinderen geboren door middel van IVF zonder cryopreservatie en kinderen geboren door middel van natuurlijke conceptie.

10


151

De resultaten van dit proefschrift tonen het belang aan van langdurige follow-up van de gezondheid en ontwikkeling van kinderen die geboren zijn uit subfertiele ouderparen, met en zonder het toepassen van ART. Het monitoren en verbeteren van de veiligheid en gezondheid van de subfertiele ouderparen en hun kinderen is noodzakelijk, met name in de huidige maatschappij waar de leeftijd van de moeder ten tijde van conceptie, subfertiliteit en het toepassen van ART toenemen en vruchtbaarheidsbehandeltechnieken continu in ontwikkeling zijn.

10

152

References

153

APPENDIX References

154

1. Centraal Bureau voor de Statistiek. Worden vrouwen steeds ouder moeder? 2014; Available at: http://www.cbs.nl/nl-NL/menu/themas/bevolking/faq/specifiek/faq-leeftijd-moeder.htm. Last accessed on July 21st, 2014.

2. Fretts RC. Effect of advanced age on fertility and pregnancy in women. 2014; Available at: http://www.uptodate.com/contents/effect-of-advanced-age-on-fertility-and-pregnancy-in-women?source=search_result&search=maternal+age&selectedTitle=1~150. Last accessed on July 21st, 2014.

3. Gindoff PR, Jewelewicz R. Reproductive potential in the older woman. Fertil Steril 1986;46:989-1001.

4. Frank O, Bianchi PG, Campana A. The end of fertility: age, fecundity and fecundability in women. J Biosoc Sci 1994;26:349-368.

5. Reproductive Endocrinology and Infertility Committee, Family Physicians Advisory Committee, Maternal-Fetal Medicine Committee, Executive and Council of the Society of Obstetricians, Liu K, Case A. Advanced reproductive age and fertility. J Obstet Gynaecol Can 2011;33:1165-1175.

6. Hansen JP. Older maternal age and pregnancy outcome: a review of the literature. Obstet Gynecol Surv 1986;41:726-742.

7. 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;15:2433-2437.

8. Steptoe PC, Edwards RG. Reimplantation of a human embryo with subsequent tubal pregnancy. Lancet 1976;1:880-882.

9. Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet 1978;2:366. 10. Ferraretti AP, Goossens V, Kupka M, Bhattacharya S, de Mouzon J, Castilla JA, et al. Assisted

reproductive technology in Europe, 2009: results generated from European registers by ESHRE. Hum Reprod 2013;28:2318-2331.

11. Kulkarni AD, Jamieson DJ, Jones HW,Jr, Kissin DM, Gallo MF, Macaluso M, et al. Fertility treatments and multiple births in the United States. N Engl J Med 2013;369:2218-2225.

12. Smeenk JMJ. Nederlandse Vereniging voor Obstetrie & Gynaecologie, IVF-cijfers 2012. 2014; Available at: http://www.nvog.nl/Sites/Files/0000003520_IVFcijfers2012.pdf. Last accessed on July 21st, 2014.

13. Palermo G, Joris H, Devroey P, Van Steirteghem AC. Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet 1992;340:17-18.

14. Zeilmaker GH, Alberda AT, van Gent I, Rijkmans CM, Drogendijk AC. Two pregnancies following transfer of intact frozen-thawed embryos. Fertil Steril 1984;42:293-296.

15. Rydhstroem H, Heraib F. Gestational duration, and fetal and infant mortality for twins vs singletons. Twin Res 2001;4:227-231.

16. Blondel B, Kogan MD, Alexander GR, Dattani N, Kramer MS, Macfarlane A, et al. The impact of the increasing number of multiple births on the rates of preterm birth and low birthweight: an international study. Am J Public Health 2002;92:1323-1330.

17. Whelan JG,3rd, Vlahos NF. The ovarian hyperstimulation syndrome. Fertil Steril 2000;73:883-896. 18. Practice Committee of American Society for Reproductive Medicine. Ovarian hyperstimulation

syndrome. Fertil Steril 2008;90:S188-93. 19. Rongieres-Bertrand C, Olivennes F, Righini C, Fanchin R, Taieb J, Hamamah S, et al. Revival of the

natural cycles in in-vitro fertilization with the use of a new gonadotrophin-releasing hormone antagonist (Cetrorelix): a pilot study with minimal stimulation. Hum Reprod 1999;14:683-688.

References

155

20. Pelinck MJ, Vogel NE, Hoek A, Simons AH, Arts EG, Mochtar MH, et al. Cumulative pregnancy rates after three cycles of minimal stimulation IVF and results according to subfertility diagnosis: a multicentre cohort study. Hum Reprod 2006;21:2375-2383.

21. Wilton L. Preimplantation genetic diagnosis for aneuploidy screening in early human embryos: a review. Prenat Diagn 2002;22:512-518.

22. Mastenbroek S, Twisk M, van Echten-Arends J, Sikkema-Raddatz B, Korevaar JC, Verhoeve HR, et al. In vitro fertilization with preimplantation genetic screening. N Engl J Med 2007;357:9-17.

23. Mastenbroek S, Twisk M, van der Veen F, Repping S. Preimplantation genetic screening: a systematic review and meta-analysis of RCTs. Hum Reprod Update 2011 ;17:454-466.

24. Harton G, Braude P, Lashwood A, Schmutzler A, Traeger-Synodinos J, Wilton L, et al. ESHRE PGD consortium best practice guidelines for organization of a PGD centre for PGD/preimplantation genetic screening. Hum Reprod 2011;26:14-24.

25. Barker DJ. Intrauterine programming of adult disease. Mol Med Today 1995;1:418-423. 26. Barker DJ. The origins of the developmental origins theory. J Intern Med 2007;261:412-417. 27. Painter RC, Roseboom TJ, Bleker OP. Prenatal exposure to the Dutch famine and disease in later

life: an overview. Reprod Toxicol 2005;20:345-352. 28. Langley-Evans SC, Welham SJ, Jackson AA. Fetal exposure to a maternal low protein diet impairs

nephrogenesis and promotes hypertension in the rat. Life Sci 1999;64:965-974. 29. Gluckman PD, Lillycrop KA, Vickers MH, Pleasants AB, Phillips ES, Beedle AS, et al. Metabolic

plasticity during mammalian development is directionally dependent on early nutritional status. Proc Natl Acad Sci U S A 2007;104:12796-12800.

30. Fleming TP, Lucas ES, Watkins AJ, Eckert JJ. Adaptive responses of the embryo to maternal diet and consequences for post-implantation development. Reprod Fertil Dev 2011;24:35-44.

31. Olivennes F, Rufat P, Andre B, Pourade A, Quiros MC, Frydman R. The increased risk of complication observed in singleton pregnancies resulting from in-vitro fertilization (IVF) does not seem to be related to the IVF method itself. Hum Reprod 1993;8:1297-1300.

32. Dumoulin JC, Land JA, Van Montfoort AP, Nelissen EC, Coonen E, Derhaag JG, et al. Effect of in vitro culture of human embryos on birthweight of newborns. Hum Reprod 2010;25:605-612.

33. Grace KS, Sinclair KD. Assisted reproductive technology, epigenetics, and long-term health: a developmental time bomb still ticking. Semin Reprod Med 2009;27:409-416.

34. Khosla S, Dean W, Reik W, Feil R. Culture of preimplantation embryos and its long-term effects on gene expression and phenotype. Hum Reprod Update 2001;7:419-427.

35. Rooke JA, McEvoy TG, Ashworth CJ, Robinson JJ, Wilmut I, Young LE, et al. Ovine fetal development is more sensitive to perturbation by the presence of serum in embryo culture before rather than after compaction. Theriogenology 2007;67:639-647.

36. Nelissen EC, Van Montfoort AP, Coonen E, Derhaag JG, Geraedts JP, Smits LJ, et al. Further evidence that culture media affect perinatal outcome: findings after transfer of fresh and cryopreserved embryos. Hum Reprod 2012;27:1966-1976.

37. Eaton JL, Lieberman ES, Stearns C, Chinchilla M, Racowsky C. Embryo culture media and neonatal birthweight following IVF. Hum Reprod 2012;27:375-379.

38. Vergouw CG, Kostelijk EH, Doejaaren E, Hompes PG, Lambalk CB, Schats R. The influence of the type of embryo culture medium on neonatal birthweight after single embryo transfer in IVF. Hum Reprod 2012;27:2619-2626.

39. Lin S, Li M, Lian Y, Chen L, Liu P. No effect of embryo culture media on birthweight and length of newborns. Hum Reprod 2013;28:1762-1767.

156

40. Bourgain C, Devroey P. The endometrium in stimulated cycles for IVF. Hum Reprod Update 2003 ;9:515-522.

41. Jovanovic A, Kramer B. The effect of hyperstimulation on transforming growth factor beta(1) and beta(2) in the rat uterus: possible consequences for embryo implantation. Fertil Steril 2010;93:1509-1517.

42. Rackow BW, Kliman HJ, Taylor HS. GnRH antagonists may affect endometrial receptivity. Fertil Steril 2008;89:1234-1239.

43. Manipalviratn S, DeCherney A, Segars J. Imprinting disorders and assisted reproductive technology. Fertil Steril 2009;91:305-315.

44. Batcheller A, Cardozo E, Maguire M, DeCherney AH, Segars JH. Are there subtle genome-wide epigenetic alterations in normal offspring conceived by assisted reproductive technologies? Fertil Steril 2011;96:1306-1311.

45. Sato A, Otsu E, Negishi H, Utsunomiya T, Arima T. Aberrant DNA methylation of imprinted loci in superovulated oocytes. Hum Reprod 2007;22:26-35.

46. Doornbos ME, Maas SM, McDonnell J, Vermeiden JP, Hennekam RC. Infertility, assisted reproduction technologies and imprinting disturbances: a Dutch study. Hum Reprod 2007;22:2476-2480.

47. Ludwig M, Katalinic A, Gross S, Sutcliffe A, Varon R, Horsthemke B. Increased prevalence of imprinting defects in patients with Angelman syndrome born to subfertile couples. J Med Genet 2005;42:289-291.

48. Landy HJ, Keith LG. The vanishing twin: a review. Hum Reprod Update 1998 Mar-Apr;4(2):177-183. 49. Mansour R, Serour G, Aboulghar M, Kamal O, Al-Inany H. The impact of vanishing fetuses on the

outcome of ICSI pregnancies. Fertil Steril 2010;94:2430-2432. 50. Pinborg A, Lidegaard O, la Cour Freiesleben N, Andersen AN. Consequences of vanishing twins in

IVF/ICSI pregnancies. Hum Reprod 2005;20:2821-2829. 51. Shebl O, Ebner T, Sommergruber M, Sir A, Tews G. Birth weight is lower for survivors of the

vanishing twin syndrome: a case-control study. Fertil Steril 2008;90:310-314. 52. 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;346:731-737. 53. 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;328:261. 54. Jackson RA, Gibson KA, Wu YW, Croughan MS. Perinatal outcomes in singletons following in vitro

fertilization: a meta-analysis. Obstet Gynecol 2004;103:551-563. 55. Sutcliffe AG, Ludwig M. Outcome of assisted reproduction. Lancet 2007;370:351-359. 56. Pandey S, Shetty A, Hamilton M, Bhattacharya S, Maheshwari A. Obstetric and perinatal

outcomes in singleton pregnancies resulting from IVF/ICSI: a systematic review and meta-analysis. Hum Reprod Update 2012;18:485-503.

57. Salem Yaniv S, Levy A, Wiznitzer A, Holcberg G, Mazor M, Sheiner E. A significant linear association exists between advanced maternal age and adverse perinatal outcome. Arch Gynecol Arch Gynecol Obstet 2011;283:755-9.

58. Tornqvist K, Finnstrom O, Kallen B, Lindam A, Nilsson E, Nygren KG, et al. Ocular malformations or poor visual acuity in children born after in vitro fertilization in Sweden. Am J Ophthalmol 2010 ;150:23-26.

59. 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;353:1746-1749.

References

157

60. Basso O, Baird DD. Infertility and preterm delivery, birthweight, and Caesarean section: a study within the Danish National Birth Cohort. Hum Reprod 2003;18:2478-2484.

61. Jaques AM, Amor DJ, Baker HW, Healy DL, Ukoumunne OC, Breheny S, et al. Adverse obstetric and perinatal outcomes in subfertile women conceiving without assisted reproductive technologies. Fertil Steril 2010;94:2674-2679.

62. Raatikainen K, Kuivasaari-Pirinen P, Hippelainen M, Heinonen S. Comparison of the pregnancy outcomes of subfertile women after infertility treatment and in naturally conceived pregnancies. Hum Reprod 2012;27:1162-1169.

63. Thomson F, Shanbhag S, Templeton A, Bhattacharya S. Obstetric outcome in women with subfertility. BJOG 2005;112:632-637.

64. Hansen M, Kurinczuk JJ, Milne E, de Klerk N, Bower C. Assisted reproductive technology and birth defects: a systematic review and meta-analysis. Hum Reprod Update 2013;19:330-353.

65. 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;366:1803-1813.

66. Zhu JL, Basso O, Obel C, Bille C, Olsen J. Infertility, infertility treatment, and congenital malformations: Danish national birth cohort. BMJ 2006;333(7570):679.

67. Yeung EH, Druschel C. Cardiometabolic health of children conceived by assisted reproductive technologies. Fertil Steril 2013;99:318-326.

68. Feuer SK, Camarano L, Rinaudo PF. ART and health: clinical outcomes and insights on molecular mechanisms from rodent studies. Mol Hum Reprod 2013;19:189-204.

69. Chen M, Norman RJ, Heilbronn LK. Does in vitro fertilisation increase type 2 diabetes and cardiovascular risk? Curr Diabetes Rev 2011;7:426-432.

70. Scherrer U, Rimoldi SF, Rexhaj E, Stuber T, Duplain H, Garcin S, et al. Systemic and pulmonary vascular dysfunction in children conceived by assisted reproductive technologies. Circulation 2012;125:1890-1896.

71. Belva F, Henriet S, Liebaers I, Van Steirteghem A, Celestin-Westreich S, Bonduelle M. Medical outcome of 8-year-old singleton ICSI children (born >or=32 weeks' gestation) and a spontaneously conceived comparison group. Hum Reprod 2007;22:506-515.

72. Belva F, Roelants M, De Schepper J, Roseboom TJ, Bonduelle M, Devroey P, et al. Blood pressure in ICSI-conceived adolescents. Hum Reprod 2012;27:3100-3108.

73. Ceelen M, van Weissenbruch MM, Roos JC, Vermeiden JP, van Leeuwen FE, Delemarre-van de Waal HA. Body composition in children and adolescents born after in vitro fertilization or spontaneous conception. J Clin Endocrinol Metab 2007;92(9):3417-3423.

74. Ceelen M, van Weissenbruch MM, Vermeiden JP, van Leeuwen FE, Delemarre-van de Waal HA. Cardiometabolic differences in children born after in vitro fertilization: follow-up study. J Clin Endocrinol Metab 2008;93:1682-1688.

75. Sakka SD, Loutradis D, Kanaka-Gantenbein C, Margeli A, Papastamataki M, Papassotiriou I, et al. Absence of insulin resistance and low-grade inflammation despite early metabolic syndrome manifestations in children born after in vitro fertilization. Fertil Steril 2010;94:1693-1699.

76. Valenzuela-Alcaraz B, Crispi F, Bijnens B, Cruz-Lemini M, Creus M, Sitges M, et al. Assisted reproductive technologies are associated with cardiovascular remodeling in utero that persists postnatally. Circulation 2013;128:1442-1450.

77. Miles HL, Hofman PL, Peek J, Harris M, Wilson D, Robinson EM, et al. In vitro fertilization improves childhood growth and metabolism. J Clin Endocrinol Metab 2007;92:3441-3445.

78. Brandes JM, Scher A, Itzkovits J, Thaler I, Sarid M, Gershoni-Baruch R. Growth and development of children conceived by in vitro fertilization. Pediatrics 1992;90:424-429.

158

79. Koivurova S, Hartikainen AL, Sovio U, Gissler M, Hemminki E, Jarvelin MR. Growth, psychomotor development and morbidity up to 3 years of age in children born after IVF. Hum Reprod 2003;18:2328-2336.

80. Bonduelle M, Bergh C, Niklasson A, Palermo GD, Wennerholm UB, Collaborative Study Group of Brussels, Gothenburg and New York. Medical follow-up study of 5-year-old ICSI children. Reprod Biomed Online 2004;9:91-101.

81. Basatemur E, Sutcliffe A. Follow-up of children born after ART. Placenta 2008;29 Suppl B:135-140. 82. Ludwig AK, Katalinic A, Thyen U, Sutcliffe AG, Diedrich K, Ludwig M. Physical health at 5.5 years

of age of term-born singletons after intracytoplasmic sperm injection: results of a prospective, controlled, single-blinded study. Fertil Steril 2009;91:115-124.

83. Ford GW, Doyle LW, Davis NM, Callanan C. Very low birth weight and growth into adolescence. Arch Pediatr Adolesc Med 2000;154:778-784.

84. Stroescu R, Micle I, Marginean O, Bizerea T, Marazan M, Puiu M, et al. Is small for gestational age status associated with an increased risk of atherogenesis? Maedica (Buchar) 2013;8:315-320.

85. Ceelen M, van Weissenbruch MM, Vermeiden JP, van Leeuwen FE, Delemarre-van de Waal HA. Pubertal development in children and adolescents born after IVF and spontaneous conception. Hum Reprod 2008;23:2791-2798.

86. Beydoun HA, Sicignano N, Beydoun MA, Matson DO, Bocca S, Stadtmauer L, et al. A cross-sectional evaluation of the first cohort of young adults conceived by in vitro fertilization in the United States. Fertil Steril 2010;94:2043-2049.

87. Belva F, Roelants M, Painter R, Bonduelle M, Devroey P, De Schepper J. Pubertal development in ICSI children. Hum Reprod 2012;27:1156-1161.

88. Mau Kai C, Main KM, Andersen AN, Loft A, Skakkebaek NE, Juul A. Reduced serum testosterone levels in infant boys conceived by intracytoplasmic sperm injection. J Clin Endocrinol Metab 2007;92:2598-2603.

89. Middelburg KJ, Heineman MJ, Bos AF, Hadders-Algra M. Neuromotor, cognitive, language and behavioural outcome in children born following IVF or ICSI-a systematic review. Hum Reprod Update 2008;14:219-231.

90. Hadders-Algra M. Two distinct forms of minor neurological dysfunction: perspectives emerging from a review of data of the Groningen Perinatal Project. Dev Med Child Neurol 2002;44:561-571.

91. Surveillance of Cerebral Palsy in Europe. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Surveillance of Cerebral Palsy in Europe (SCPE). Dev Med Child Neurol 2000;42:816-824.

92. Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M, Damiano D, et al. A report: the definition and classification of cerebral palsy April 2006. Dev Med Child Neurol Suppl 2007;109:8-14.

93. Hvidtjorn D, Schieve L, Schendel D, Jacobsson B, Svaerke C, Thorsen P. Cerebral palsy, autism spectrum disorders, and developmental delay in children born after assisted conception: a systematic review and meta-analysis. Arch Pediatr Adolesc Med 2009;163:72-83.

94. Hvidtjorn D, Grove J, Schendel D, Svaerke C, Schieve LA, Uldall P, et al. Multiplicity and early gestational age contribute to an increased risk of cerebral palsy from assisted conception: a population-based cohort study. Hum Reprod 2010;25:2115-2123.

95. Zhu JL, Hvidtjorn D, Basso O, Obel C, Thorsen P, Uldall P, et al. Parental infertility and cerebral palsy in children. Hum Reprod 2010;25:3142-3145.

96. Sutcliffe AG, Taylor B, Li J, Thornton S, Grudzinskas JG, Lieberman BA. Children born after intracytoplasmic sperm injection: population control study. BMJ 1999;318:704-705.

References

159

97. Zhu JL, Basso O, Obel C, Hvidtjorn D, Olsen J. Infertility, infertility treatment and psychomotor development: the Danish National Birth Cohort. Paediatr Perinat Epidemiol 2009;23:98-106.

98. Zhu JL, Obel C, Basso O, Olsen J. Parental infertility and developmental coordination disorder in children. Hum Reprod 2010;25:908-913.

99. Morin NC, Wirth FH, Johnson DH, Frank LM, Presburg HJ, Van de Water VL, et al. Congenital malformations and psychosocial development in children conceived by in vitro fertilization. J Pediatr 1989;115:222-227.

100. Levy-Shiff R, Vakil E, Dimitrovsky L, Abramovitz M, Shahar N, Har-Even D, et al. Medical, cognitive, emotional, and behavioral outcomes in school-age children conceived by in-vitro fertilization. J Clin Child Psychol 1998;27:320-329.

101. Sutcliffe AG, Taylor B, Saunders K, Thornton S, Lieberman BA, Grudzinskas JG. Outcome in the second year of life after in-vitro fertilisation by intracytoplasmic sperm injection: a UK case-control study. Lancet 2001;357:2080-2084.

102. Agarwal P, Loh SK, Lim SB, Sriram B, Daniel ML, Yeo SH, et al. Two-year neurodevelopmental outcome in children conceived by intracytoplasmic sperm injection: prospective cohort study. BJOG 2005;112:1376-1383.

103. Wikstrand MH, Stromland K, Flodin S, Bergh C, Wennerholm UB, Hellstrom A. Ophthalmological findings in children born after intracytoplasmic sperm injection. Acta Ophthalmol Scand 2006;84:177-181.

104. Knoester M, Vandenbroucke JP, Helmerhorst FM, van der Westerlaken LA, Walther FJ, Veen S. Matched follow-up study of 5-8 year old ICSI-singletons: comparison of their neuromotor development to IVF and naturally conceived singletons. Hum Reprod 2007;22:1638-1646.

105. Ludwig A, Katalinic A, Thyen U, Sutcliffe AG, Diedrich K, Ludwig M. Neuromotor development and mental health at 5.5 years of age of singletons born at term after intracytoplasmatic sperm injection ICSI: results of a prospective controlled single-blinded study in Germany. Fertil Steril 2009;91:125-132.

106. Ludwig AK, Hansen A, Katalinic A, Sutcliffe AG, Diedrich K, Ludwig M, et al. Assessment of vision and hearing in children conceived spontaneously and by ICSI: a prospective controlled, single-blinded follow-up study. Reprod Biomed Online 2010;20:391-397.

107. Leunens L, Celestin-Westreich S, Bonduelle M, Liebaers I, Ponjaert-Kristoffersen I. Cognitive and motor development of 8-year-old children born after ICSI compared to spontaneously conceived children. Hum Reprod 2006;21:2922-2929.

108. Leunens L, Celestin-Westreich S, Bonduelle M, Liebaers I, Ponjaert-Kristoffersen I. Follow-up of cognitive and motor development of 10-year-old singleton children born after ICSI compared with spontaneously conceived children. Hum Reprod 2008;23:105-11.c

109. Leslie GI, Gibson FL, McMahon C, Cohen J, Saunders DM, Tennant C. Children conceived using ICSI do not have an increased risk of delayed mental development at 5 years of age. Hum Reprod 2003;18:2067-2072.

110. Wagenaar K, Ceelen M, van Weissenbruch MM, Knol DL, Delemarre-van de Waal HA, Huisman J. School functioning in 8- to 18-year-old children born after in vitro fertilization. Eur J Pediatr 2008;167:1289-1295.

111. Wagenaar K, van Weissenbruch MM, Knol DL, Cohen-Kettenis PT, Delemarre-van de Waal HA, Huisman J. Information processing, attention and visual-motor function of adolescents born after in vitro fertilization compared with spontaneous conception. Hum Reprod 2009;24:913-921.

112. Mains L, Zimmerman M, Blaine J, Stegmann B, Sparks A, Ansley T, et al. Achievement test performance in children conceived by IVF. Hum Reprod 2010;25:2605-2611.

160

113. Knoester M, Helmerhorst FM, Vandenbroucke JP, van der Westerlaken LA, Walther FJ, Veen S, et al. Cognitive development of singletons born after intracytoplasmic sperm injection compared with in vitro fertilization and natural conception. Fertil Steril 2008;90:289-296.

114. Goldbeck L, Gagsteiger F, Mindermann I, Strobele S, Izat Y. Cognitive development of singletons conceived by intracytoplasmic sperm injection or in vitro fertilization at age 5 and 10 years. J Pediatr Psychol 2009;34:774-781.

115. Gucuyener K, Arhan E, Soysal AS, Ergenekon E, Turan O, Onal E, et al. Language assessment of non-handicapped twins at 5 years of age. Pediatr Int 2011;53:944-949.

116. Carson C, Kelly Y, Kurinczuk JJ, Sacker A, Redshaw M, Quigley MA. Effect of pregnancy planning and fertility treatment on cognitive outcomes in children at ages 3 and 5: longitudinal cohort study. BMJ 2011;343:d4473.

117. Hvidtjorn D, Grove J, Schendel D, Schieve LA, Svaerke C, Ernst E, et al. Risk of autism spectrum disorders in children born after assisted conception: a population-based follow-up study. J Epidemiol Community Health 2011;65:497-502.

118. Bhutta AT, Cleves MA, Casey PH, Cradock MM, Anand KJ. Cognitive and behavioral outcomes of school-aged children who were born preterm: a meta-analysis. JAMA 2002 Aug 14;288(6):728-737.

119. Mick E, Biederman J, Prince J, Fischer MJ, Faraone SV. Impact of low birth weight on attention-deficit hyperactivity disorder. J Dev Behav Pediatr 2002;23:16-22.

120. Wagenaar K, Huisman J, Cohen-Kettenis PT, Delemarre-van de Waal HA. An overview of studies on early development, cognition, and psychosocial well-being in children born after in vitro fertilization. J Dev Behav Pediatr 2008;29:219-230.

121. Knoester M, Helmerhorst FM, van der Westerlaken LA, Walther FJ, Veen S, Leiden Artificial Reproductive Techniques Follow-up Project (L-art-FUP). Matched follow-up study of 5 8-year-old ICSI singletons: child behaviour, parenting stress and child (health-related) quality of life. Hum Reprod 2007;22:3098-3107.

122. Wagenaar K, van Weissenbruch MM, van Leeuwen FE, Cohen-Kettenis PT, Delemarre-van de Waal HA, Schats R, et al. Self-reported behavioral and socioemotional functioning of 11- to 18-year-old adolescents conceived by in vitro fertilization. Fertil Steril 2011;95:611-616.

123. Berry KA, Baron IS, Weiss BA, Baker R, Ahronovich MD, Litman FR. In vitro fertilization and late preterm preschoolers' neuropsychological outcomes: the PETIT study. Am J Obstet Gynecol 2013;209:356.e1-356.e6.

124. Carson C, Redshaw M, Sacker A, Kelly Y, Kurinczuk JJ, Quigley MA. Effects of pregnancy planning, fertility, and assisted reproductive treatment on child behavioral problems at 5 and 7 years: evidence from the Millennium Cohort Study. Fertil Steril 2013;99:456-463.

125. Sandin S, Nygren KG, Iliadou A, Hultman CM, Reichenberg A. Autism and mental retardation among offspring born after in vitro fertilization. JAMA 2013;310:75-84.

126. Wagenaar K, van Weissenbruch MM, Knol DL, Cohen-Kettenis PT, Delemarre-van de Waal HA, Huisman J. Behavior and socioemotional functioning in 9-18-year-old children born after in vitro fertilization. Fertil Steril 2009;92:1907-1914.

127. Zachor DA, Ben Itzchak E. Assisted reproductive technology and risk for autism spectrum disorder. Res Dev Disabil 2011;32:2950-2956.

128. Anderson V, Northam E, Hendy J, Wrennall J. Chapter 2: Cerebral development. Developmental Neuropsychology - A Clinical Approach. 1st edition ed.: Psychology Press, Taylor & Francis Group, Hove and New York, USA; 2001. p. 39-58.

129. Kalat JW. Chapter 5: Development and Plasticity of the Brain. Biological Pshychology. 9th edition ed.: Thomson Wadsworth, Belmont, USA; 2007. p. 121-136.

References

161

130. Volpe JJ. Unit I: Human brain development. Neurology of the Newborn. 5th edition ed.: Elsevier Health, Oxford, UK; 2013. p. 1-51.

131. de Graaf-Peters VB, Hadders-Algra M. Ontogeny of the human central nervous system: what is happening when? Early Hum Dev 2006;82:257-266.

132. Hadders-Algra M. Early brain damage and the development of motor behavior in children: clues for therapeutic intervention? Neural Plast 2001;8:31-49.

133. Lunsing RJ, Hadders-Algra M, Huisjes HJ, Touwen BC. Minor neurological dysfunction from birth to 12 years. II: Puberty is related to decreased dysfunction. Dev Med Child Neurol 1992;34:404-409.

134. Herting MM, Maxwell EC, Irvine C, Nagel BJ. The impact of sex, puberty, and hormones on white matter microstructure in adolescents. Cereb Cortex 2012;22:1979-1992.

135. Middelburg KJ, Haadsma ML, Heineman MJ, Bos AF, Hadders-Algra M. Ovarian hyperstimulation and the in vitro fertilization procedure do not influence early neuromotor development; a history of subfertility does. Fertil Steril 2010;93:544-553.

136. Pelinck MJ, Vogel NE, Arts EG, Simons AH, Heineman MJ, Hoek A. Cumulative pregnancy rates after a maximum of nine cycles of modified natural cycle IVF and analysis of patient drop-out: a cohort study. Hum Reprod 2007;22:2463-2470.

137. Pelinck MJ, Knol HM, Vogel NE, Arts EG, Simons AH, Heineman MJ, et al. Cumulative pregnancy rates after sequential treatment with modified natural cycle IVF followed by IVF with controlled ovarian stimulation. Hum Reprod 2008;23:1808-1814.

138. Van der Steeg JW, Steures P, Eijkemans MJ, Habbema JD, Hompes PG, Broekmans FJ, et al. Pregnancy is predictable: a large-scale prospective external validation of the prediction of spontaneous pregnancy in subfertile couples. Hum Reprod 2007;22:536-542.

139. Van Loendersloot LL, van Wely M, Limpens J, Bossuyt PM, Repping S, van der Veen F. Predictive factors in in vitro fertilization (IVF): a systematic review and meta-analysis. Hum Reprod Update 2010;16:577-589.

140. Middelburg KJ, Heineman MJ, Bos AF, Pereboom M, Fidler V, Hadders-Algra M. The Groningen ART cohort study: ovarian hyperstimulation and the in vitro procedure do not affect neurological outcome in infancy. Hum Reprod 2009;24:3119-3126.

141. Middelburg KJ, Jan Heineman M, Haadsma ML, Bos AF, Kok JH, Hadders-Algra M. Neurological condition of infants born after in vitro fertilization with preimplantation genetic screening. Pediatr Res 2010;67:430-434.

142. Middelburg KJ, van der Heide M, Houtzager B, Jongbloed-Pereboom M, Fidler V, Bos AF, et al. Mental, psychomotor, neurologic, and behavioral outcomes of 2-year-old children born after preimplantation genetic screening: follow-up of a randomized controlled trial. Fertil Steril 2011;96:165-169.

143. Heineman KR, Bos AF, Hadders-Algra M. The Infant Motor Profile: a standardized and qualitative method to assess motor behaviour in infancy. Dev Med Child Neurol 2008;50:275-282.

144. Edelman GM. Neural Darwinism: The Theory of Neuronal group Selection.. Oxford, The United Kingdom: Oxford University press; 1989.

145. Hadders-Algra M. Variation and variability: key words in human motor development. Phys Ther 2010;90:1823-1837.

146. Heineman KR, Hadders-Algra M. Evaluation of neuromotor function in infancy-A systematic review of available methods. J Dev Behav Pediatr 2008;29:315-323.

147. Hadders-Algra M, Brogren E, Katz-Salamon M, Forssberg H. Periventricular leucomalacia and preterm birth have different detrimental effects on postural adjustments. Brain 1999;122:727-740.

162

148. Hadders-Algra M. Putative neural substrate of normal and abnormal general movements. Neurosci Biobehav Rev 2007;31:1181-1190.

149. Spittle AJ, Brown NC, Doyle LW, Boyd RN, Hunt RW, Bear M, et al. Quality of general movements is related to white matter pathology in very preterm infants. Pediatrics 2008;121:e1184-9.

150. Hadders-Algra M. Reduced variability in motor behaviour: an indicator of impaired cerebral connectivity? Early Hum Dev 2008;84:787-789.

151. Heineman KR, Bos AF, Hadders-Algra M. Infant Motor Profile and cerebral palsy: promising associations. Dev Med Child Neurol 2011;53Suppl4:40-45.

152. Butcher PR, van Braeckel K, Bouma A, Einspieler C, Stremmelaar EF, Bos AF. The quality of preterm infants' spontaneous movements: an early indicator of intelligence and behaviour at school age. J Child Psychol Psychiatry 2009;50:920-930.

153. Hielkema T, Blauw-Hospers CH, Dirks T, Drijver-Messelink M, Bos AF, Hadders-Algra M. Does physiotherapeutic intervention affect motor outcome in high-risk infants? An approach combining a randomized controlled trial and process evaluation. Dev Med Child Neurol 2011;53:e8-15.

154. Heineman KR, La Bastide-Van Gemert S, Fidler V, Middelburg KJ, Bos AF, Hadders-Algra M. Construct validity of the Infant Motor Profile: relation with prenatal, perinatal, and neonatal risk factors. Dev Med Child Neurol 2010;52:e209-15.

155. Hadders-Algra M. The neuromotor examination of the preschool child and its prognostic significance. Ment Retard Dev Disabil Res Rev 2005;11:180-188.

156. Hadders-Algra M. Developmental coordination disorder: is clumsy motor behavior caused by a lesion of the brain at early age? Neural Plast 2003;10:39-50.

157. Batstra L, Neeleman J, Hadders-Algra M. The neurology of learning and behavioural problems in pre-adolescent children. Acta Psychiatr Scand 2003 Aug;108:92-100.

158. Peters LH, Maathuis CG, Hadders-Algra M. Limited motor performance and minor neurological dysfunction at school age. Acta Paediatr 2011;100:271-278.

159. Punt M, DE Jong M, DE Groot E, Hadders-Algra M. Minor neurological dysfunction in children with dyslexia. Dev Med Child Neurol 2010;52:1127-1132.

160. Prechtl HF. The optimality concept. Early Hum Dev 1980;4:201-205. 161. Huisman M, Koopman-Esseboom C, Lanting CI, van der Paauw CG, Tuinstra LG, Fidler V, et al.

Neurological condition in 18-month-old children perinatally exposed to polychlorinated biphenyls and dioxins. Early Hum Dev 1995;43:165-176.

162. Hadders-Algra M, Mavinkurve-Groothuis AM, Groen SE, Stremmelaar EF, Martijn A, Butcher PR. Quality of general movements and the development of minor neurological dysfunction at toddler and school age. Clin Rehabil 2004;18:287-299.

163. Lanting CI, Patandin S, Weisglas-Kuperus N, Touwen BC, Boersma ER. Breastfeeding and neurological outcome at 42 months. Acta Paediatr 1998;87:1224-1229.

164. Bouwstra H, Dijck-Brouwer J, Decsi T, Boehm G, Boersma ER, Muskiet FA, et al. Neurologic condition of healthy term infants at 18 months: positive association with venous umbilical DHA status and negative association with umbilical trans-fatty acids. Pediatr Res 2006;60:334-339.

165. Escolano-Margarit MV, Ramos R, Beyer J, Csabi G, Parrilla-Roure M, Cruz F, et al. Prenatal DHA status and neurological outcome in children at age 5.5 years are positively associated. J Nutr 2011;141:1216-1223.

166. Kaufman AS, Kaufman NL. Kaufman Assessment Battery for Children, Second edition. 2nd ed. Circle Pines, MN: AGS Publishing; 2004.

References

163

167. Achenbach TM. Manual for the Child Behavior Checklist 1½-5 and 1991 Profile. Burlington, VT; 1991.

168. Verhulst FC, Van der Ende J, Koot HM. Handleiding voor de CBCL/4-18. Rotterdam, The Netherlands: Sophia Kinderziekenhuis, Erasmus Medisch Centrum; 2000.

169. Bonduelle M, Wennerholm UB, Loft A, Tarlatzis BC, Peters C, Henriet S, et al. A multi-centre cohort study of the physical health of 5-year-old children conceived after intracytoplasmic sperm injection, in vitro fertilization and natural conception. Hum Reprod 2005;20:413-419.

170. Ceelen M, van Weissenbruch MM, Prein J, Smit JJ, Vermeiden JP, Spreeuwenberg M, et al. Growth during infancy and early childhood in relation to blood pressure and body fat measures at age 8-18 years of IVF children and spontaneously conceived controls born to subfertile parents. Hum Reprod 2009;24:2788-2795.

171. Griesinger G, Kolibianakis EM, Diedrich K, Ludwig M. Ovarian stimulation for IVF has no quantitative association with birthweight: a registry study. Hum Reprod 2008;23:2549-2554.

172. Schendelaar P, Middelburg KJ, Bos AF, Heineman MJ, Jongbloed-Pereboom M, Hadders-Algra M. The Groningen ART cohort study: the effects of ovarian hyperstimulation and the IVF laboratory procedures on neurological condition at 2 years. Hum Reprod 2011;26:703-712.

173. Van den Heuvel ER. A comparison of estimation methods on the coverage probability of Satterthwaite confidence intervals for assay precision with unbalanced data. Commun Stat-Simul 2010;39:777-794.

174. Cioni G, Ferrari F, Einspieler C, Paolicelli PB, Barbani MT, Prechtl HF. Comparison between observation of spontaneous movements and neurologic examination in preterm infants. J Pediatr 1997;130:704-711.

175. 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;24:2158-2172.

176. Pinborg A, Loft A, Aaris Henningsen AK, Rasmussen S, Andersen AN. Infant outcome of 957 singletons born after frozen embryo replacement: the Danish National Cohort Study 1995-2006. Fertil Steril 2010;94:1320-1327.

177. Steel AJ, Sutcliffe A. Long-term health implications for children conceived by IVF/ICSI. Hum Fertil (Camb) 2009;12:21-27.

178. Diggle PJ, Liang KY, Zeger SL. Design considerations. Analysis of Longitudinal Data. first ed. Oxford, United Kingdom: Oxford University Press; 1994. p. 31.

179. Pinborg A, Loft A, Schmidt L, Andersen AN. Morbidity in a Danish national cohort of 472 IVF/ICSI twins, 1132 non-IVF/ICSI twins and 634 IVF/ICSI singletons: health-related and social implications for the children and their families. Hum Reprod 2003;18:1234-1243.

180. 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;85:1586-1622.

181. Nargund G, Fauser BC, Macklon NS, Ombelet W, Nygren K, Frydman R, et al. The ISMAAR proposal on terminology for ovarian stimulation for IVF. Hum Reprod 2007;22:2801-2804.

182. Hempel MS. The Neurological Examination for Toddler-Age. Groningen: University of Groningen; 1993.

183. Hadders-Algra M. The examination of the child with minor neurological dysfunction. 3rd ed. London: Mac Keith Press; 2010.

164

184. Carson C, Kurinczuk JJ, Sacker A, Kelly Y, Klemetti R, Redshaw M, et al. Cognitive development following ART: effect of choice of comparison group, confounding and mediating factors. Hum Reprod 2010;25:244-252.

185. Ludwig AK, Katalinic A, Entenmann A, Thyen U, Sutcliffe AG, Diedrich K, et al. Can we sense ART? The blinded examiner is not blind-a problem with follow-up studies on children born after assisted reproduction. Fertil Steril 2009;92:950-952.

186. 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;14:1250-1254.

187. 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;88:823-829.

188. Bouwstra H, Boersma ER, Boehm G, Dijck-Brouwer DA, Muskiet FA, Hadders-Algra M. Exclusive breastfeeding of healthy term infants for at least 6 weeks improves neurological condition. J Nutr 2003;133:4243-4245.

189. Bouwstra H, Dijck-Brouwer DA, Boehm G, Boersma ER, Muskiet FA, Hadders-Algra M. Long-chain polyunsaturated fatty acids and neurological developmental outcome at 18 months in healthy term infants. Acta Paediatr 2005;94:26-32.

190. de Jong C, Kikkert HK, Fidler V, Hadders-Algra M. Effects of long-chain polyunsaturated fatty acid supplementation of infant formula on cognition and behaviour at 9 years of age. Dev Med Child Neurol 2012;54:1102-1108.

191. Seggers J, Schendelaar P, Bos AF, Heineman MJ, Middelburg KJ, Haadsma ML, et al. Increased time to pregnancy is associated with suboptimal neurological condition of 2-year-olds. Arch Dis Child Fetal Neonatal Ed 2013;98:F434-6.

192. Bowen JR, Gibson FL, Leslie GI, Saunders DM. Medical and developmental outcome at 1 year for children conceived by intracytoplasmic sperm injection. Lancet 1998;351:1529-1534.

193. De Jong M, Punt M, De Groot E, Minderaa RB, Hadders-Algra M. Minor neurological dysfunction in children with autism spectrum disorder. Dev Med Child Neurol 2011;53:641-646.

194. Bay B, Mortensen EL, Kesmodel US. Assisted reproduction and child neurodevelopmental outcomes: a systematic review. Fertil Steril 2013;100:844-853.

195. Conti E, Mazzotti S, Calderoni S, Saviozzi I, Guzzetta A. Are children born after assisted reproductive technology at increased risk of autism spectrum disorders? A systematic review. Hum Reprod 2013;28:3316-3327.

196. Groen SE, de Blecourt AC, Postema K, Hadders-Algra M. General movements in early infancy predict neuromotor development at 9 to 12 years of age. Dev Med Child Neurol 2005;47:731-738.

197. Giedd JN, Castellanos FX, Rajapakse JC, Vaituzis AC, Rapoport JL. Sexual dimorphism of the developing human brain. Prog Neuropsychopharmacol Biol Psychiatry 1997;21:1185-1201.

198. Lenroot RK, Gogtay N, Greenstein DK, Wells EM, Wallace GL, Clasen LS, et al. Sexual dimorphism of brain developmental trajectories during childhood and adolescence. Neuroimage 2007;36:1065-1073.

199. Kikkert HK, de Jong C, Hadders-Algra M. Minor neurological dysfunction and cognition in 9-year-olds born at term. Early Hum Dev 2013;89:263-270.

200. Kikkert HK, de Jong C, van den Heuvel ER, Hadders-Algra M. Minor neurological dysfunction and behaviour in 9-year-old children born at term: evidence for sex dimorphism. Dev Med Child Neurol 2013;55:1023-1029.

References

165

201. Jongbloed-Pereboom M, Middelburg KJ, Heineman MJ, Bos AF, Haadsma ML, Hadders-Algra M. The impact of IVF/ICSI on parental well-being and anxiety 1 year after childbirth. Hum Reprod 2012;27:2389-95.

202. Berg BJ, Wilson JF. Psychological functioning across stages of treatment for infertility. J Behav Med 1991;14:11-26.

203. Ulrich D, Gagel DE, Hemmerling A, Pastor VS, Kentenich H. Couples becoming parents: something special after IVF? J Psychosom Obstet Gynaecol 2004;25:99-113.

204. Sarkar P, Bergman K, O'Connor TG, Glover V. Maternal antenatal anxiety and amniotic fluid cortisol and testosterone: possible implications for foetal programming. J Neuroendocrinol 2008;20:489-496.

205. Talge NM, Neal C, Glover V, Early Stress, Translational Research and Prevention Science Network: Fetal and Neonatal Experience on Child and Adolescent Mental Health. Antenatal maternal stress and long-term effects on child neurodevelopment: how and why? J Child Psychol Psychiatry 2007;48:245-261.

206. Dahlof LG, Hard E, Larsson K. Influence of maternal stress on offspring sexual behaviour. Anim Behav 1977;25:958-968.

207. Dahlof LG, Hard E, Larsson K. Influence of maternal stress on the development of the fetal genital system. Physiol Behav 1978;20:193-195.

208. Kinsley CH, Bridges RS. Prenatal stress and maternal behavior in intact virgin rats: response latencies are decreased in males and increased in females. Horm Behav;22:76-89.

209. Ward IL, Stehm KE. Prenatal stress feminizes juvenile play patterns in male rats. Physiol Behav 1991;50:601-605.

210. Wu MV, Shah NM. Control of masculinization of the brain and behavior. Curr Opin Neurobiol 2011;21:116-123.

211. Barrett ES, Parlett LE, Sathyanarayana S, Liu F, Redmon JB, Wang C, et al. Prenatal exposure to stressful life events is associated with masculinized anogenital distance (AGD) in female infants. Physiol Behav 2013;114-115:14-20.

212. Barrett ES, Redmon JB, Wang C, Sparks A, Swan SH. Exposure to prenatal life events stress is associated with masculinized play behavior in girls. Neurotoxicology 2014;41:20-27.

213. Zhu JL, Obel C, Basso O, Henriksen TB, Bech BH, Hvidtjorn D, et al. Infertility, infertility treatment and behavioural problems in the offspring. Paediatr Perinat Epidemiol 2011;25:466-477.

214. Zhu JL, Obel C, Basso O, Bech BH, Henriksen TB, Olsen J. Infertility, infertility treatment, and mixed-handedness in children. Early Hum Dev 2009;85:745-749.

215. Schendelaar P, Heineman KR, Heineman MJ, Jongbloed-Pereboom M, La Bastide-Van Gemert S, Middelburg KJ, et al. Movement variation in infants born following IVF/ICSI with and without ovarian hyperstimulation. Early Hum Dev 2013;89:507-513.

216. Schendelaar P, Van der Heuvel ER, Heineman MJ, La Bastide-Van Gemert S, Middelburg KJ, Seggers J. Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not. Hum Reprod 2014;29:2773-86.

217. La Bastide-Van Gemert S, Seggers J, Haadsma ML, Heineman MJ, Middelburg KJ, Roseboom TJ, et al. Is ovarian hyperstimulation associated with higher blood pressure in 4-year-old IVF offspring? Part II: an explorative causal inference approach. Hum Reprod 2014;29:510-517.

218. Hadders-Algra M, Touwen BC. Minor neurological dysfunction is more closely related to learning difficulties than to behavioral problems. J Learn Disabil 1992;25:649-657.

219. Pearl J. Causality: Models, Reasoning, and Inference. Cambridge, UK: Cambridge University Press; 2009. p. 1-165.

166

220. La Bastide-van Gemert S, Stolk RP, van den Heuvel ER, Fidler V. Causal inference algorithms can be useful in life course epidemiology. J Clin Epidemiol 2014;67:190-198.

221. Pearl J. Causal inference in statistics: an overview. Stat Surv 2009:3:96-146. 222. Spirtes P, Glymour C, Scheines R. Causation, Prediction, and Search. Cambridge, MA, USA: The

MIT Press; 2000. p. 1-123. 223. Bollen KA. Structural equations with latent variables. New York, NY etc.: Wiley; 1989. 224. Fox J. Structural equation modeling with the sem package in R. Struct Equ Modeling 2006:13:465-

486. 225. The Tetrad Project. Causal models and statistical data. Available at:

http://www.phil.cmu.edu/projects/tetrad. Pittsburgh. 2010. Last accessed on April 4th, 2014 226. R Development Core Team. R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria. Available at: http://www.R-project.org/. Last accessed on April 4th, 2014

227. Korb KB, Wallace CS. In search of the philosopher’s stone: remarks on Humphreys and Freedman’s critique of causal discovery. Br J Philos Sci 1997;48:543-553.

228. Nekkebroeck J, Van den Broeck W, Desmyttere S, Ponjaert-Kristoffersen I, Bonduelle M. The mental, motor, socio-emotional and language development of 2-year-old twins born after PGD/PGS and parental well-being. Hum Reprod 2012;27:299-301.

229. Banerjee I, Shevlin M, Taranissi M, Thornhill A, Abdalla H, Ozturk O, et al. Health of children conceived after preimplantation genetic diagnosis: a preliminary outcome study. Reprod Biomed Online 2008 Mar;16:376-381.

230. Baird DD, Wilcox AJ, Kramer MS. Why might infertile couples have problem pregnancies? Lancet 1999 May 22;353:1724-1725.

231. Liebaers I, Desmyttere S, Verpoest W, De Rycke M, Staessen C, Sermon K, et al. Report on a consecutive series of 581 children born after blastomere biopsy for preimplantation genetic diagnosis. Hum Reprod 2010;25:275-282.

232. Nekkebroeck J, Bonduelle M, Desmyttere S, Van den Broeck W, Ponjaert-Kristoffersen I. Socio-emotional and language development of 2-year-old children born after PGD/PGS, and parental well-being. Hum Reprod 2008;23:1849-1857.

233. Nekkebroeck J, Bonduelle M, Desmyttere S, Van den Broeck W, Ponjaert-Kristoffersen I. Mental and psychomotor development of 2-year-old children born after preimplantation genetic diagnosis/screening. Hum Reprod 2008;23:1560-1566.

234. Desmyttere S, De Schepper J, Nekkebroeck J, De Vos A, De Rycke M, Staessen C, et al. Two-year auxological and medical outcome of singletons born after embryo biopsy applied in preimplantation genetic diagnosis or preimplantation genetic screening. Hum Reprod 2009;24:470-476.

235. Desmyttere S, De Rycke M, Staessen C, Liebaers I, De Schrijver F, Verpoest W, et al. Neonatal follow-up of 995 consecutively born children after embryo biopsy for PGD. Hum Reprod 2012;27:288-293.

236. De Mouzon J, Goossens V, Bhattacharya S, Castilla JA, Ferraretti AP, Korsak V, et al. Assisted reproductive technology in Europe, 2006: results generated from European registers by ESHRE. Hum Reprod 2010;25:1851-1862.

237. Chapman DA, Scott KG, Mason CA. Early risk factors for mental retardation: role of maternal age and maternal education. Am J Ment Retard 2002;107:46-59.

238. Harper JC, Wilton L, Traeger-Synodinos J, Goossens V, Moutou C, SenGupta SB, et al. The ESHRE PGD Consortium: 10 years of data collection. Hum Reprod Update 2012;18:234-247.

References

167

239. Pandian Z, Bhattacharya S, Ozturk O, Serour G, Templeton A. Number of embryos for transfer following in-vitro fertilisation or intra-cytoplasmic sperm injection. Cochrane Database Syst Rev 2009;7:CD003416.

240. Min JK, Hughes E, Young D, Gysler M, Hemmings R, Cheung AP, et al. Elective single embryo transfer following in vitro fertilization. J Obstet Gynaecol Can 2010;32:363-377.

241. Sullivan EA, Wang YA, Hayward I, Chambers GM, Illingworth P, McBain J, et al. Single embryo transfer reduces the risk of perinatal mortality, a population study. Hum Reprod 2012;27:3609-3615.

242. Jongbloed-Pereboom M, Middelburg KJ, Heineman MJ, Bos AF, Hadders-Algra M. The Groningen assisted reproductive technologies cohort study: developmental status and behavior at 2 years. Fertil Steril 2011;95:2283-2289.

243. Seggers J, Haadsma ML, Bastide-van Gemert S, Heineman MJ, Kok JH, Middelburg KJ, et al. Blood pressure and anthropometrics of 4-y-old children born after preimplantation genetic screening: follow-up of a unique, moderately sized, randomized controlled trial. Pediatr Res 2013;74:606-614.

244. Plachot M. Chromosomal abnormalities in oocytes. Mol Cell Endocrinol 2001;183Suppl1:S59-63. 245. Ertzeid G, Storeng R. The impact of ovarian stimulation on implantation and fetal development in

mice. Hum Reprod 2001;16:221-225. 246. Valbuena D, Martin J, de Pablo JL, Remohi J, Pellicer A, Simon C. Increasing levels of estradiol are

deleterious to embryonic implantation because they directly affect the embryo. Fertil Steril 2001;76:962-968.

247. Kondapalli LA, Perales-Puchalt A. Low birth weight: is it related to assisted reproductive technology or underlying infertility? Fertil Steril 2013;99:303-310.

248. Pelinck MJ, Hadders-Algra M, Haadsma ML, Nijhuis WL, Kiewiet SM, Hoek A, et al. Is the birthweight of singletons born after IVF reduced by ovarian stimulation or by IVF laboratory procedures? Reprod Biomed Online 2010;21:245-251.

249. Jongmans M, Mercuri E, de Vries L, Dubowitz L, Henderson SE. Minor neurological signs and perceptual-motor difficulties in prematurely born children. Arch Dis Child Fetal Neonatal Ed 1997;76:F9-14.

250. Soorani-Lunsing RJ, Hadders-Algra M, Olinga AA, Huisjes HJ, Touwen BC. Is minor neurological dysfunction at 12 years related to behaviour and cognition? Dev Med Child Neurol 1993;35:321-330.

251. van Hoorn JF, Maathuis CG, Peters LH, Hadders-Algra M. Handwriting, visuomotor integration, and neurological condition at school age. Dev Med Child Neurol 2010;52:941-947.

252. Kikkert HK, DE Jong C, Hadders-Algra M. Minor neurological dysfunction and IQ in 9-year-old children born at term. Dev Med Child Neurol 2011;53:e16-25.

253. Peters LH, Maathuis KG, Kouw E, Hamming M, Hadders-Algra M. Test-retest, inter-assessor and intra-assessor reliability of the modified Touwen examination. Eur J Paediatr Neurol 2008;12:328-333.

254. Macklon NS, Geraedts JP, Fauser BC. Conception to ongoing pregnancy: the 'black box' of early pregnancy loss. Hum Reprod Update 2002;8:333-343.

255. Magli MC, Jones GM, Gras L, Gianaroli L, Korman I, Trounson AO. Chromosome mosaicism in day 3 aneuploid embryos that develop to morphologically normal blastocysts in vitro. Hum Reprod 2000;15:1781-1786.

256. Baart EB, Martini E, van den Berg I, Macklon NS, Galjaard RJ, Fauser BC, et al. Preimplantation genetic screening reveals a high incidence of aneuploidy and mosaicism in embryos from young women undergoing IVF. Hum Reprod 2006;21:223-233.

168

257. Vanneste E, Voet T, Le Caignec C, Ampe M, Konings P, Melotte C, et al. Chromosome instability is common in human cleavage-stage embryos. Nat Med 2009;15:577-583.

258. Van Echten-Arends J, Mastenbroek S, Sikkema-Raddatz B, Korevaar JC, Heineman MJ, van der Veen F, et al. Chromosomal mosaicism in human preimplantation embryos: a systematic review. Hum Reprod Update 2011;17:620-627.

259. Munne S, Alikani M, Tomkin G, Grifo J, Cohen J. Embryo morphology, developmental rates, and maternal age are correlated with chromosome abnormalities. Fertil Steril 1995;64:382-391.

260. Coonen E, Derhaag JG, Dumoulin JC, van Wissen LC, Bras M, Janssen M, et al. Anaphase lagging mainly explains chromosomal mosaicism in human preimplantation embryos. Hum Reprod 2004;19:316-324.

261. Baart EB, Martini E, Eijkemans MJ, Van Opstal D, Beckers NG, Verhoeff A, et al. Milder ovarian stimulation for in-vitro fertilization reduces aneuploidy in the human preimplantation embryo: a randomized controlled trial. Hum Reprod 2007;22:980-988.

262. Buck Louis GM, Schisterman EF, Dukic VM, Schieve LA. Research hurdles complicating the analysis of infertility treatment and child health. Hum Reprod 2005;20:12-18.

263. Bensdorp AJ, Slappendel E, Koks C, Oosterhuis J, Hoek A, Hompes P, et al. The INeS study: prevention of multiple pregnancies: a randomised controlled trial comparing IUI COH versus IVF e SET versus MNC IVF in couples with unexplained or mild male subfertility. BMC Womens Health 200918;9:35-6874-9-35.

264. Carson C, Sacker A, Kelly Y, Redshaw M, Kurinczuk JJ, Quigley MA. Asthma in children born after infertility treatment: findings from the UK Millennium Cohort Study. Hum Reprod 2013;28:471-479.

265. Hargreave M, Jensen A, Toender A, Andersen KK, Kjaer SK. Fertility treatment and childhood cancer risk: a systematic meta-analysis. Fertil Steril 2013;100:150-161.

266. Hart R, Norman RJ. The longer-term health outcomes for children born as a result of IVF treatment: Part I--General health outcomes. Hum Reprod Update 2013;19:232-243.

267. Kallen B. The risk of neurodisability and other long-term outcomes for infants born following ART. Semin Fetal Neonatal Med 2014;19:239-244.

268. Practice Committee of Society for Assisted Reproductive Technology, Practice Committee of American Society for Reproductive Medicine. Elective single-embryo transfer. Fertil Steril 2012;97:835-842.

269. Pelkonen S, Koivunen R, Gissler M, Nuojua-Huttunen S, Suikkari AM, Hyden-Granskog C, et al. Perinatal outcome of children born after frozen and fresh embryo transfer: the Finnish cohort study 1995-2006. Hum Reprod 2010;25:914-923.

270. Pinborg A, Henningsen AA, Loft A, Malchau SS, Forman J, Andersen AN. Large baby syndrome in singletons born after frozen embryo transfer (FET): is it due to maternal factors or the cryotechnique? Hum Reprod 2014;29:618-627.

271. Pinborg A. To transfer fresh or thawed embryos? Semin Reprod Med 2012 Jun;30(3):230-235. 272. Henningsen AK, Pinborg A. Birth and perinatal outcomes and complications for babies conceived

following ART. Semin Fetal Neonatal Med 2014;19:234-238. 273. Wells D, Alfarawati S, Fragouli E. Use of comprehensive chromosomal screening for embryo

assessment: microarrays and CGH. Mol Hum Reprod 2008;14:703-710. 274. Hall S. Louise Brown, first test tube baby, is pregnant. The Guardian. July 11, 2006; Available at:

http://www.theguardian.com/science/2006/jul/11/medicineandhealth.lifeandhealth. Last accessed on September 30th, 2014.

References

169

275. BBC News. Baby son joy for test-tube mother. 2007, January 14; Available at: http://news.bbc.co.uk/2/hi/uk_news/6260171.stm. Last accessed on September 30th, 2014.

276. De Schepper J, Belva F, Schiettecatte J, Anckaert E, Tournaye H, Bonduelle M. Testicular growth and tubular function in prepubertal boys conceived by intracytoplasmic sperm injection. Horm Res 2009;71:359-363.

277. Belva F, Bonduelle M, Painter RC, Schiettecatte J, Devroey P, De Schepper J. Serum inhibin B concentrations in pubertal boys conceived by ICSI: first results. Hum Reprod 2010;25:2811-2814.

278. Belva F, Bonduelle M, Schiettecatte J, Tournaye H, Painter RC, Devroey P, et al. Salivary testosterone concentrations in pubertal ICSI boys compared with spontaneously conceived boys. Hum Reprod 2011;26:438-441.

170

171

APPENDIX Dankwoord (Acknowledgements)

172

Promoveren doe je niet alleen. Dit proefschrift was er nooit geweest zonder de hulp, steun en inspanningen van vele mensen. Ik ben dan ook mijn dank verschuldigd aan een ieder die op welke manier dan ook heeft bijgedragen aan de totstandkoming van dit proefschrift. In het bijzonder wil ik de volgende mensen bedanken, in de eerste plaats mijn promotores prof. dr. M. Hadders-Algra en prof. dr. M. J. Heineman en copromotor dr. K.J. Middelburg. Mijna, bedankt dat je me hebt meegenomen naar wetenschapsland. Wetenschapsland, zo noem jij dat, een toch wat simpele bewoording voor iets dat zo moeilijk, groot en vaak abstract is. Maar dat is nu ook precies wat ik aan jou bewonder: hoe jij, door het geheel te overzien, iets ingewikkelds weet te reduceren tot iets simpels. Ja, je lijkt altijd alles te begrijpen en te weten. Ik ben blij dat ik me heb mogen voeden met jouw kennis en deskundigheid. Je bent niet alleen mijn rolmodel geweest qua analytisch vermogen en wetenschappelijk redeneren, maar ook qua persoon. Mijna, ik vind jou een powervrouw: hoe jij succesvol een afdeling runt, razendsnel op mails reageert, een hoogleraarschap combineert met een universitaire opleiding, soms letterlijk in één dag de halve wereld overvliegt en nooit klaagt: ik bewonder jou. Je inspireert mij een succesvolle carrière na te streven en mijn doelen te bereiken. Bedankt. Maas Jan, mijn promotor op afstand, letterlijk misschien, figuurlijk was je namelijk dichtbij. Regelmatig verruilde je het AMC voor een bezoekje aan het Hoge Noorden om zo de IVF-vergaderingen bij te wonen, waarvoor mijn dank en waardering. Ik bewonder je warmte, enthousiasme en relativerend vermogen. Deze eigenschappen en jouw positiviteit hebben mij geholpen het promotieproces te doorlopen en keuzes te maken in het leven. Je introduceerde mij het dertigers dilemma, de ‘quarterlife’ crisis: de oneindige hoeveelheid keuzemogelijkheden in vrijwel elk aspect van het leven waar je halverwege de twintig mee geconfronteerd wordt zijn natuurlijk niet de minsten. Bedankt voor je betrokkenheid in dezen. Bedankt ook dat je, ondanks al je activiteiten, altijd tijdig mijn vragen wist te beantwoorden, manuscripten van feedback wist te voorzien en, ook niet onbelangrijk, altijd vliegensvlug de author signature forms retourneerde. Ik ben er trots op dat ik onder jouw begeleiding promoveer. Het leven van een AIOS gaat niet altijd over rozen. Karin, jij zult dit beamen. Ik bewonder je uithoudingsvermogen: tussen je drukke bestaan als gynaecoloog in opleiding en moeder van twee door, heb je het aangedurfd copromotor te worden en tijd gevonden om mij te begeleiden bij mijn onderzoek. Die tijd moest je zo nu en dan wel even zoeken. Maar als je die gevonden had resulteerde dat voor mij in effectiviteit en inzichtelijkheid. We hebben samen veel overlegd, allerlei knopen doorgehakt en je hebt mij ontzettend goed geholpen met wetenschappelijk schrijven. Ik heb jou, als ervaringsdeskundige, eindeloos vragen mogen stellen over ‘het promoveren’ en ‘de verdere carrière’. Daarnaast stond je letterlijk aan mijn zijde tijdens een zenuwslopende presentatie op het NVOG-congres. Ik ben vereerd dat ik jouw onderzoek heb mogen voortzetten. Jij was het die mij wees op de mogelijkheid van een MD/PhD-plek op het IVF-project. Bedankt voor deze mooie kans, je betrokkenheid en begeleiding. De leden van de beoordelingscommissie, prof. dr. O.F. Brouwer, prof. dr. S.A. Scherjon en prof dr. F. van der Veen wil ik heel hartelijk bedanken voor de kritische blik op mijn manuscript en het positieve eindoordeel van dit proefschrift.

Dankwoord (Acknowledgements)

173

De Junior Scientific Masterclass wil ik bedanken voor de mogelijkheid die ik heb gekregen promotieonderzoek te combineren met de master Geneeskunde in de vorm van een MD/PhD-traject. Klinische onderzoeken zijn afhankelijk van proefpersonen. Ik wil alle ouders en hun kinderen dan ook bedanken voor hun deelname aan de onderzoeken in dit proefschrift. Geheel vrijwillig en door nieuwsgierigheid gedreven hebben zij bijgedragen aan dit proefschrift, waarvoor mijn volle waardering. Veel dank ben ik verschuldigd aan iedereen die heeft meegeschreven aan dit proefschrift: mijn co-auteurs. Prof. dr. A.F. Bos, beste Arie, samen met Mijna en Maas Jan heb jij het IVF-project op de kaart gezet. Bedankt voor jouw hulp in het verwezenlijken van mijn MD/PhD-traject, wat geresulteerd heeft in mooie publicaties en uiteindelijk dit proefschrift. Jouw klinische invalshoek is van toegevoegde waarde geweest voor onze artikelen. Ik kijk uit naar de verdediging van mijn proefschrift tegenover de oppositie, waarin jij deelneemt. Prof. dr. J.H. Kok wil ik bedanken voor de prettige samenwerking binnen de PGS-studie. Prof. dr. E.R. van den Heuvel en dr. S. la Bastide-van Gemert, beste Edwin en Sacha, bedankt voor jullie rekenhulp bij onze klinisch logische, maar statistisch vaak uitdagende vraagstukken. Jullie hebben, jawel, significant bijgedragen aan dit proefschrift en mijn wetenschappelijke ontwikkeling. Drs. M. Jongbloed-Pereboom en dr. K.R. Heineman, Marjolein en Kirsten. Jullie waren al lang bedreven onderzoekers op de afdeling Ontwikkelingsneurologie toen ik er voor het eerst binnenstapte. Een goede voorbereiding is het halve werk en daarin heb jij, Marjolein, voor mij een belangrijke rol gespeeld. Dankzij jou ben ik vertrouwd geraakt met het onderzoeken van kleine kinderen, iets waar ik nauwelijks kaas van gegeten had. Ik heb SPSS ook pas echt leren begrijpen nadat ik met jou achter de pc had gezeten. Kirsten, jij hebt mij veel over de IMP geleerd en mij scherp en kritisch gehouden bij het beoordelen van alle filmpjes. Bedankt voor de prettige samenwerking en mooie bijdrage aan dit proefschrift. Als onderzoeker kom je vaak handen te kort en wat is het dan prettig dat er enthousiaste studenten Geneeskunde en Psychologie een stage wetenschap op onze afdeling komen doen. Rosan Aapkes, Elise Bennink, Hanneke Geut, Inonge Reimert, Aïsha Sinot, Jolien Snijders en Bertine de Vries, jullie liepen stage op het IVF-project. Ik ben jullie zeer dankbaar voor al jullie hulp bij het onderzoeken van de jonge kinderen en het verwerken van de onderzoeksgegevens. De kinderen uit dit proefschrift zijn natuurlijk al lang geen peuters meer. Inmiddels is de follow-up op de leeftijd van 9 jaar in volle gang. Mijn dank gaat uit naar alle studenten die hebben meegeholpen of momenteel meehelpen met deze follow-up ronde: An Bennema, Manuelle Dania, Lysanne Jansma, Susanne van Kessel, Derk Kuiper, Thi Yen Nguyen en Elon Reijrink, bedankt! Dr. M.L. Haadsma, beste Maaike, jij bent vanaf het begin al betrokken bij het IVF-project. In het prille begin deed jij de inclusie van de kinderen, werkte je samen met Karin en ben je uiteindelijk copromotor geworden van Jorien. Hoewel je nooit direct betrokken was bij mijn onderzoek, heb ik toch erg veel aan jouw warme en bemoedigende persoonlijkheid gehad. Je was regelmatig op ons ‘kantoor’ te

174

vinden met oplossingen, wijze woorden en dat hart onder de riem. Ik zal ook ons ESHRE congresavontuur in Istanbul niet vergeten, wat hebben we een indrukken opgedaan. Maaike, bedankt voor je investering in het IVF-project en je persoonlijke betrokkenheid. Een prettige werkplek en leuke collega’s maken het leven van een onderzoeker erg aangenaam. Lieve en leuke (oud-)onderzoekers van de afdeling Ontwikkelingsneurologie. Mattana (May) Angsupaisal, Lieke van Balen, Anke Boxum, Tineke Dirks, Ilse Ebbers-Dekkers, Elisa Hamer, Tjitske Hielkema, Siebrig Hooijsma, Corina de Jong, Hedwig Kikkert, Lieke Peters en Rivka Toonen, ik wil jullie bedanken voor de heerlijke tijd. Bedankt voor alle gezelligheid, de koffie, het eindeloos geklets, maar vooral ook jullie betrokkenheid en steun. Linze Jaap Dijkstra en Michiel Schrier, jullie hebben met jullie technisch oog meerdere malen weten te voorkomen dat ik mijn pc door het raam heb gegooid, waarvoor dank. Anneke Kracht en Loes de Weerd, lieve Anneke en Loes, bedankt voor al jullie logistieke en administratieve hulp, maar vooral ook voor jullie luisterend oor en alle pret. Graag bedank ik Rinze Schols voor zijn hulp bij het ontwerp en de lay-out van de omslag. Rinze, ik ken jou van een bijbaantje in de kledingwinkel, waar we altijd al spraken over onze passie: tekenen. Ik wilde vroeger naar de kunstacademie, jij gíng vroeger naar de kunstacademie en je ontwikkelde je tot grafisch vormgever. Ik vind het erg speciaal dat onze wegen elkaar opnieuw kruisten en dat we ons teken-gepraat hebben omgezet in daad. Thanks, de omslag had niet mooier kunnen zijn! Mijn paranimfen, Jorien Seggers en Anne van Zoonen, mijn beider steun en toeverlaat, wat ben ik vereerd met jullie aan mijn zijde. Jorien, mijn collega en lieve vriendin, mijn partner in crime. Van keuzeproject neurochirurgie, via een proefproject vol kinderen met ADHD, naar MD/PhD-ers op het IVF-project. En nu is het zover, beide als promovenda onze boekjes verdedigen, maar ook beide als elkaars paranimf voor support. Dit typeert ons ook: we zijn met en door elkaar gegroeid en we hebben het elkaar altijd gegund. Ik bewonder je gedrevenheid, je enorme doorzettingsvermogen en onuitputtelijke wilskracht. Ik heb hier zo nu en dan van kunnen profiteren door me af en toe een beetje aan je op te trekken. We hebben samen veel gelachen, goed Engels geleerd, de nodige frustraties doorgemaakt, hypotheses gegenereerd, klassen kinderen onderzocht, op (inter)nationale congressen gesproken, oneindig geanalyseerd, maar vooral een prachtige, innige vriendschap opgebouwd. Jor, ik had me deze dag niet beter kunnen voorstellen. Ik ben dan ook dankbaar en blij dat onze wegen zich hebben mogen kruisen. En.. ik ben trots op je! Lieve Anne, studiegenoot van het eerste uur en dierbare vriendin. Wij hebben samen de studie Geneeskunde doorlopen. De vele studie-uren in de bibliotheek, maar natuurlijk met name de gezelligheid, etentjes, drankjes en een aantal bizarre feestjes hebben bijgedragen aan een leuke studententijd. Studie, toekomst, twijfels, sport.. jij stond altijd voor mij klaar met een luisterend oor en een goed advies. Anne, met jou kan ik lief-en-leed delen, jij hebt zo’n groot hart. Ik bewonder ook je doorzettingskracht; als jij iets wilt, dan ga je ervoor en bereik je dat ook. Ik ben blij dat ik jou heb kunnen aansteken met het MD/PhD-virus en dat ik over een paar maanden getuige mag zijn van jouw promotie. Bedankt voor alle moeite die jij in vandaag hebt gestopt, bedankt voor je goede raad, je aanmoedigingen en onze mooie vriendschap. Hoewel niet direct betrokken, maar minstens zo belangrijk voor ál die andere dingen in het leven, immers: ‘You will never feel truly satisfied by work until you are satisfied by life’ (Heather Schuck,

Dankwoord (Acknowledgements)

175

Amerikaanse schrijfster). Vrienden van school, studie en coschappen, huisgenoten, vrienden van karate, de vele bijbaantjes en andere mooie ontmoetingen, bedankt! Mijn waardering en dank gaat ook uit naar mijn schoonfamilie, in het bijzonder mijn schoonouders Imo Grolleman en Sijke van der Heide. Wij kennen elkaar alweer zo lang dat jullie mij van de Geneeskunde loting tot de buluitreiking hebben meegemaakt. Wat ben ik blij dat jullie nu ook bij mijn verdediging aanwezig zijn. Bedankt voor jullie warmte en jullie steun en interesse in mijn onderzoek. Mijn grote, ‘kleine’ broer, Brian-Roy Schendelaar. Lieve Brian-Roy, wij zijn als water en vuur, maar dan in de positieve zin van het woord. Hoe fijn is het voor mij geweest om bij jou even los te komen van al dat rekenen en geschrijf. Bedankt voor de mentale support, bedankt dat je zo trots op mij bent. Wist je trouwens dat vanuit ontwikkelingsneurologisch perspectief het tweede kind neurologisch optimaler is dan het eerste kind (alle andere factoren buitenbeschouwing gelaten)? Laat jij nu toevallig de jongste zijn.. Dit proefschrift draag ik op aan de vier mensen in mijn leven die steeds weer de drijvende kracht zijn in alles wat ik doe: mijn ouders, Harrie en Hellen Schendelaar, mijn oma Krisnawatie Soechitram en mijn tante Henna Gadjradj. Lieve Oma, lieve tante Henna, allebei zijn jullie veel te vroeg uit mijn leven verdwenen, maar jullie geloof in en liefde voor mij zal ik altijd bij me dragen, bedankt daarvoor. Lieve Paps en Mams, als kind riep ik al dat ik dokter wilde worden, iets dat jullie altijd hebben aangemoedigd. Dat de vroege ontwikkeling medebepalend is voor je latere leven, zoals ik min of meer in dit proefschrift beweer, wisten jullie kennelijk al lang. Bedankt voor jullie onvoorwaardelijke liefde, toewijding en stimulans, het heeft me hier gebracht. Tot slot, Tom Grolleman. Mijn Tom en de allerliefste, inmiddels zelf expert op het gebied van dit proefschrift dankzij het eindeloos aanhoren van mijn oefenpresentaties en het checken van mijn manuscripten op spelfouten. Bedankt in de ruimste zin van het woord. Bedankt voor je liefde en begrip, je onvoorwaardelijke steun, het oppeppen, (regelmatig) helpen relativeren en bijspringen wanneer dat nodig was. Jouw auto werd onze auto toen ik Noord-Nederland door moest om kinderen te onderzoeken. En jij had nu eenmaal een veel praktischer camerastatief (sorry dat hij vervolgens een jaar lang op de afdeling heeft gestaan en jij je maar afvroeg waar hij was). En, hoe vaak je die zware koffer wel niet naar de afdeling gebracht hebt? Duizend maal dank voor wie jij voor mij bent. Ik kijk uit naar ons volgende avontuur en de rest van ons leven.

176

Over de auteur (About the author)

177

APPENDIX Over de auteur (About the author)

178

Pamela Schendelaar werd op 2 december 1987 geboren in Amersfoort. In haar kinderjaren verhuisde zij naar Leeuwarden, waar zij in 2006 haar middelbare schooldiploma behaalde aan het Christelijk Gymnasium Beyers Naudé. In datzelfde jaar startte zij met de studie Geneeskunde aan de Rijksuniversiteit Groningen. In de bachelorfase werd haar interesse voor onderzoek gewekt, wat resulteerde in het succesvol doorlopen van het Bachelor Honours Traject van de Junior Scientific Masterclass (JSM). De ambitie om te promoveren groeide, waarop zij in 2010 haar Master Geneeskunde startte met de stage wetenschap bij prof. dr. M. Hadders-Algra op de afdeling Ontwikkelingsneurologie van het Universitair Medisch Centrum Groningen (UMCG). Na deze stage solliciteerde zij naar een promotieplaats in de vorm van een MD/PhD-traject. Dit traject biedt studenten Geneeskunde de mogelijkheid om de masterfase van de studie te combineren met promotieonderzoek. In 2011 startte Pamela met het MD/PhD-traject onder begeleiding van prof. dr. M. Hadders-Algra, prof. dr. M.J. Heineman en dr. K.J.Middelburg. Zij combineerde haar coschappen in het UMCG en het Medisch Centrum Leeuwarden (MCL) met het doen van onderzoek en deed haar oudste coschap, de semi-artsstage, bij de afdeling Obstetrie & Gynaecologie in het UMCG. In februari start Pamela als arts-assistent niet in opleiding tot specialist op de afdeling Heelkunde in het MCL. Uiteindelijk wil zij zich graag specialiseren in de plastische chirurgie, het specialisme dat haar hart van de gynaecologie gestolen heeft.

List of publications

179

APPENDIX List of publications

180

Schendelaar P, Middelburg KJ, Bos AF, Heineman MJ, Jongbloed-Pereboom M, Hadders-Algra M. The Groningen ART cohort study: the effects of ovarian hyperstimulation and the IVF laboratory procedures on neurological condition at 2 years. Hum Reprod 2011;26:703-12. Seggers J, Schendelaar P, Bos AF, Heineman MJ, Middelburg KJ, Haadsma ML, Hadders-Algra M. Increased time to pregnancy is associated with suboptimal neurological condition of 2-year-olds. Arch Dis Child Fetal Neonatal Ed 2013;98:F434-6. Schendelaar P, Middelburg KJ, Bos AF, Heineman MJ, Kok JH, La Bastide-Van Gemert S, Seggers J, Van den Heuvel ER, Hadders-Algra M. The effect of preimplantation genetic screening on neurological, cognitive and behavioural development in 4-year-old children: follow-up of a RCT. Hum Reprod 2013;28:1508-18. Schendelaar P, Heineman KR, Heineman MJ, Jongbloed-Pereboom M, La Bastide-Van Gemert S, Middelburg KJ, Van den Heuvel ER, Hadders-Algra M. Movement variation in infants born following IVF/ICSI with and without ovarian hyperstimulation. Early Hum Dev 2013;89:507-13. Seggers J, Haadsma ML, Bastide-van Gemert Sl, Heineman MJ, Kok JH, Middelburg KJ, Roseboom TJ, Schendelaar P, Van den Heuvel ER, Hadders-Algra M. Blood pressure and anthropometrics of 4-y-old children born after preimplantation genetic screening: follow-up of a unique, moderately sized, randomized controlled trial. Pediatr Res 2013;74:606-14. Seggers J, Haadsma ML, La Bastide-Van Gemert S, Heineman MJ, Middelburg KJ, Roseboom TJ, Schendelaar P, Van den Heuvel ER, Hadders-Algra M. Is ovarian hyperstimulation associated with higher blood pressure in 4-year-old IVF offspring? Part I: multivariable regression analysis. Hum Reprod 2014;29:502-9. La Bastide-Van Gemert S, Seggers J, Haadsma ML, Heineman MJ, Middelburg KJ, Roseboom TJ, Schendelaar P, Hadders-Algra M, Van den Heuvel ER. Is ovarian hyperstimulation associated with higher blood pressure in 4-year-old IVF offspring? Part II: an explorative causal inference approach. Hum Reprod 2014;29:510-7. Schendelaar P, Van den Heuvel ER, Heineman MJ, La Bastide-Van Gemert S, Middelburg KJ, Seggers J, Hadders-Algra M. Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not. Hum Reprod 2014;29:2773-86. Schendelaar P, la Bastide-van Gemert S, Heineman MJ, Middelburg KJ, Seggers J, van den Heuvel ER, Hadders-Algra M. Subfertility factors rather than assisted conception factors affect cognitive and behavioral development of 4-year-old singletons. Submitted.

Documents

University of Groningen Offspring of subfertile couples ... · Cover design: Rinze Schols, drawings by Rinze Schols & Pamela Schendelaar, 2014 . Thesis lay-out: Pamela Schendelaar