University of Groningen
Offspring of subfertile couples: neurodevelopmental outcome at preschool age
Schendelaar, Pamela
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher'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.]
Copyright
Other 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 the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 18-06-2017
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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
Edwin R. van den Heuvel
Maas Jan Heineman
Sacha la Bastide-van Gemert
Karin J. Middelburg
Jorien Seggers
Mijna Hadders-Algra
Human Reproduction 2014 ; 29 : 2773 – 2786
67
ABSTRACT
5
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 4year-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.
68
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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
69
INTRODUCTION
5
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 longterm 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 7year-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
70
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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 COHIVF 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
71
5
5
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
72
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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.
73
5
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
5
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 MNCIVF 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).
74
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
TABLE I. Characteristics of participating parents and children of the ART study groups.
COH - IVF
n = 63
Characteristics
MNC - IVF
n = 53
Sub - NC
n = 79
Child characteristics
Male gender, n (%)
34 (54.0)
26 (49.1)
41 (51.9)
First born, n (%)
43 (68.3)
37 (69.8)
49 (62.0)
50 (47.5 - 60.1)
48.9 (48.0 - 52.5)
48.9 (47.9 - 56.4)
39.4 (33.4 - 42.3)*
40.1 (34.6 - 42.6)
40.0 (30.1 - 42.7)*
7 (11.1)
6 (11.3)
5 (6.3)
3393.1 (563.2)*
3384.4 (585.7)*
3577.9 (519.4)*/*
Corrected age at examination at 4 years of age
(months), median (range)
Birth characteristics
Gestational age (weeks), median (range)
Preterm birth (< 37 weeks), n (%)
Birthweight (grams), mean (SD)
Low birthweight (< 2500 gram), n (%)
Small for gestational agea, n (%)
3 (4.8)
4 (7.5)
3 (3.8)
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)
1 (1.3)
Neonatal characteristics
Apgar score 5 min < 7c, n (%)
Neonatal intensive care admission, n (%)
Breastfed for > 6 weeksc, n (%)
0
0
1 (1.6)
2 (3.8)
5 (6.3)
28 (46.7)
24 (45.3)
39 (50.0)
Parental characteristics
Maternal 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 parameters
Time 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.
8 (12.7)*/**
1 (1.9)*
0**
Vanishing twins, n (%)
f
Subfertility causes
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)***/***
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.
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.
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.
f
Couples may have more than one cause of subfertility, therefore totals may exceed 100% .
* P < 0.05; ** P < 0.01; *** P < 0.001.
75
5
5
FIGURE I. Flow chart of the Groningen ART cohort.
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
76
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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).
77
5
78
4 (13.8)
10 (34.5)
0
5 (14.7)
11 (32.4)
0
Simple MND, n (%)
Complex MND, n (%)
13 (44.8)
0
0
Dysfunctional reflexes, n (%)
Visuomotor dysfunction, n (%)
0
25 (31.6)
19 (30.2)
17 (27.0)
0
0
21 (33.3)
9 (14.3)
33 (52.4)
47.6 [46.7 ; 48.4]
12.0 [11.6 ; 12.4]
total
n = 63
0
13 (50.0)
3 (11.5)
5 (19.2)
0
0
3 (11.5)
4 (15.4)
19 (73.1)
48.2 [46.8 ; 49.5]
12.0 [11.5 ; 12.6]
boys
n = 26
0
9 (33.3)
6 (22.2)
7 (25.9)
0
0
5 (18.5)
7 (25.9)
15 (55.6)
49.0 [47.7 ; 50.3]
12.7 [12.2 ; 13.3]
girls
n = 27
MNC-IVF
0
19 (24.1)
9 (17.0)
12 (22.6)
0
0
8 (15.1)
11 (20.8)
34 (64.2)
48.6 [47.6 ; 49.5]
12.4 [12.0 ; 12.8]
total
n = 53
0
15 (36.6)
10 (24.4)
14 (34.1)
0
0
7 (17.1)
11 (26.8)
23 (56.1)
47.6 [46.5 ; 48.6]
11.8 [11.3 ; 12.2]
boys
n = 41
0
14 (36.8)
9 (23.7)
11 (28.9)
0
0
11 (28.9)
6 (15.8)
21 (55.3)
49.1 [48.0 ; 50.3]
12.5 [12.0 ; 12.9]
girls
n = 38
Sub-NC
0
29 (36.7)
19 (24.1)
25 (31.6)
0
0
18 (22.8)
17 (21.5)
44 (55.7)
48.4 [47.4 ; 49.2]
12.1 [11.8 ; 12.4]
total
n = 79
17 (18.1)
26 (27.7)
0
20 (19.8)
21 (20.8)
0
Simple MND, n (%)
Complex MND, n (%)
25 (26.6)
36 (38.3)
0
22 (21.8)
45 (44.6)
0
Posture and muscle tone dysfunction, n (%)
Dysfunctional reflexes, n (%)
Visuomotor dysfunction, n (%)
0
81 (41.5)
47 (24.1)
54 (27.7)
0 (0)
0
47 (24.1)
37 (19.0)
111 (56.9)
48.2 [47.7 ; 48.6]
12.2 [11.9 ; 12.4]
total
n =195
0
22 (40.7)
10 (18.5)
16 (29.6)
0
0
10 (18.5)
12 (22.2)
32 (59.3)
48.6 [47.6 ; 49.5]
12.5 [12.1 ; 12.9]
boys
n = 54
0
16 (36.4)
8 (18.2)
4 (9.1)
0
0
4 (9.1)
23 (16.7)
34 (77.3)
50.2 [49.2 ; 51.2]
13.1 [12.7 ; 13.6]
Reference groupb
girls
n = 44
0
38 (38.8)
18 (18.4)
20 (20.4)
0 (0)
0
14 (14.3)
18 (18.4)
66 (67.3)
49.4 [48.7 ; 50.1]
12.8 [12.5 ; 13.1]
total
n = 98
Subfertile group: COH-IVF, MNC-IVF and Sub-NC groups together.
Reference group: naturally conceived children born to fertile parents.
a
b
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. NOS = neurological
optimality score, MND = minor neurological dysfunction.
0
24 (25.5)
0
30 (29.7)
Fine motor dysfunction, n (%)
Gross motor dysfunction, n (%)
Domains of neurological dysfunction
Neurologically abnormal, n (%)
51 (54.3)
60 (59.4)
Neurologically normal, n (%)
Neurological classification
12.5 [12.2 ; 12.8]
48.8 [48.1 ; 49.5]
11.8 [11.5 ; 12.1]
47.5 [46.9 ; 48.2]
Fluency score, mean [95% CI]
Subfertile groupa
girls
n = 94
Neurological optimality score, mean [95% CI]
Neurological optimality
boys
n = 101
TABLE III. Neurological optimality, neurological classification and domains of neurological dysfunction of the subfertile and reference 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. NOS = neurological
optimality score, MND = minor neurological dysfunction.
10 (34.5)
9 (26.5)
17 (50.0)
Posture and muscle tone dysfunction, n (%)
0
6 (20.7)
0
11 (32.4)
Fine motor dysfunction, n (%)
Gross motor dysfunction, n (%)
Domains of neurological dysfunction
Neurologically abnormal, n (%)
15 (51.7)
18 (52.9)
Neurologically normal, n (%)
Neurological classification
12.3 [11.7 ; 12.8]
48.1 [46.8 ; 49.4]
11.7 [11.2 ; 12.2]
47.0 [45.8 ; 48.2]
Fluency score, mean [95% CI]
girls
n = 29
COH-IVF
Neurological optimality score, mean [95% CI]
Neurological optimality
boys
n = 34
TABLE II. Neurological optimality, neurological classification and domains of neurological dysfunction of the ART study groups.
5
0.431 [0.067 ; 2.761]
6.217 [1.052 ; 2.930]
MNC-IVF versus Sub-NCa
COH-IVF versus MNC-IVFa
1.000 [0.801 ; 1.247]
Time to pregnancyd (years)e
0.998
0.923
0.044
0.374
0.089
P -value
0.528
0.439
0.759
0.339
P -value
0.752
0.322 [-0.501 ; 1.145]
-0.143 [-1.064 ; 0.779]
-0.306 [-0.937 ; 0.325]
Mean estimate [95% CI]
0.024 [-0.127 ; 0.176]
0.070
0.003
0.321
0.675 [0.311 ; 1.467]
1.542 [1.161 ; 2.047]
P -value
Adjusted odds ratio [95% CI]
3.139 [0.920 ; 10.824]
0.349 [0.120 ; 1.018]
0.172
-0.240 [-0.587 ; 0.106]
0.059
P -value
Mean estimate [95% CI]
0.225
0.176
-0.963 [-2.363 ; 0.438]
0.199 [0.037 ; 1.061]
0.644
-0.492 [-2.603 ; 1.619]
2.824 [0.527 ; 15.131]
1.132 [0.734 ; 1.743]
0.936
0.077 [-1.810 ; 1.963]
1.207 [1.038 ; 1.404]
1.524 [0.699 ; 3.325]
2.934 [1.028 ; 8.370]
Adjusted odds ratio [95% CI]
-0.039 [-0.276 ; 0.197]
Mean estimate [95% CI]
-0.877 [-0.914 ; 0.160]
-0.887 [-2.319 ; 0.546]
0.340 [-0.985 ; 1.664]
0.689
Adjusted mean difference [95%
CI]
-0.547 [-1.910 ; 0.816]
0.081 [-0.019 ; 0.180]
Mean estimate [95% CI]
-0.383 [-0.827 ; 0.061]
-0.258 [-0.861 ; 0.346]
Adjusted mean difference [95%
CI]
-0.415 [-2.472 ; 1.642]
P -value
0.692
0.216 [-0.342 ; 0.773]
Adjusted mean difference [95%
CI]
-0.042 [-0.616 ; 0.532]
Adjusted mean difference [95%
CI]
0.179 [-0.719 ; 1.077]
P -value
TOTAL
GIRLS
0. 014
0.289
0.044
0.054
0.575
P -value
0.744
P -value
0.097
0.223
0.614
0.429
P -value
0.110
P -value
0.090
0.400
0.447
0.885
P -value
0.034
0.132
0.232
P -value
0.222
P -value
0.683
0.819
P -value
0.347
P -value
0.879
0.963
P -value
interaction
effect with sex
Adjusted for gestational age, birthweight, vanishing twins, time to pregnancy, parental age and parental educational level.
Subfertile group: COH-IVF, MNC-IVF and Sub-NC groups together.
Adjusted for gestational age, birthweight, vanishing twins, Caesarean section, breastfed for > 6 weeks, firstborn, parental age and parental educational level.
Only time to pregnancy data of couples of the subfertile group were used for the analyses.
Adjusted for gestational age, birthweight, vanishing twins, parental age and parental educational level.
a
b
c
d
e
Note that in general P < 0.05 were considered to be significant. Bonferroni corrections were applied in the analyses on pairwise comparisons of the three Groningen ART cohort study groups, P < 0.05/3 = 0.017 were considered to be significant.
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. NOS = neurological optimality score, MND = minor neurological dysfunction.
0.946 [0.305 ; 2.930]
Subfertile groupb versus Reference groupc
Time to pregancyd (years)
1.798 [0.914 ; 3.536]
Adjusted odds ratio [95% CI]
COH-IVF versus Sub-NCa
Outcome measure: complex MND
0.102 [-0.218 ; 0.422]
P -value
Mean estimate [95% CI]
Time to pregnancyd (years)e
0.354
-0.720 [-2.251 ; 0.812]
0.379
-0.909 [-2.953 ; 1.134]
Subfertile groupb versus Reference groupc
0.677
0.604
P -value
COH-IVF versus MNC-IVFa
Adjusted mean difference [95%
CI]
-0.509 [-2.456 ; 1.437]
0.400 [-1.504 ; 2.304]
a
0.095
MNC-IVF versus Sub-NCa
COH-IVF versus Sub-NC
Outcome measure: NOS
0.110 [-0.019 ; 0.239]
P -value
Mean estimate [95% CI]
Time to pregnancyd (years)e
0.222
-0.389 [-1.015 ; 0.237]
Subfertile groupb versus Reference groupc
0.762
-0.127 [-0.954 ; 0.701]
COH-IVF versus MNC-IVFa
0.931
0.815
P -value
0.034 [-0.738 ; 0.805]
Adjusted mean difference [95%
CI]
-0.093 [-0.881 ; 0.695]
MNC-IVF versus Sub-NCa
COH-IVF versus Sub-NC
a
Outcome measure: FLUENCY SCORE
BOYS
TABLE IV. Outcomes of the effect of aspects of assisted conception and the presence and severity of subfertility on the neurological outcome of 4-year-olds.
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
5
79
TABLE V. Prevalence of complex MND per parental TTP category.
No complex MND
Complex MND
Total groupa
Boysa
Girls
Total group
Boys
Girls
n = 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%)
MND: minor neurological dysfunction, TTP: time to pregnancy
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).
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.
5
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 nonparticipants (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
80
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
TABLE VI. Characteristics of parents and children of the subfertile group and the reference group.
Subfertile group
n = 195
Characteristics
Reference group
n = 98
P -value
Child characteristics
Male gender, n (%)
101 (51.8)
54 (55.1)
0.621
First born, n (%)
129 (66.2)
46 (46.9)
0.002
48.9 (47.5 - 60.1)
49.1 (48.0 - 54.6)
0.217
40.0 (30.1 - 42.7)
40.1 (32.0 - 42.4)
0.432
18 (9.2)
3 (3.1)
0.058
3465.6 (557.2)
3599.7 (507.0)
0.049
10 (5.1)
2 (2.1)
0.348
Corrected age at examination at 4 years of age
(months), median (range)
Birth characteristics
Gestational age (weeks)a, median (range)
Preterm birth (< 37 weeks)a, n (%)
Birthweight (grams)a, mean (sd)
Low birthweight (< 2500 gram)a, n (%)
Small for gestational agea,b, n (%)
4 (2.1)
4 (4.3)
0.281
Caesarean section, n (%)
44 (22.6)
9 (9.2)
0.006
Signs of fetal distressc, n (%)
68 (34.9)
29 (29.6)
0.430
Neonatal characteristics
Apgar score 5 min < 7a, n (%)
1 (0.5)
3 (4.2)
0.062
Neonatal intensive care admission, n (%)
8 (4.1)
10 (10.2)
0.067
91 (47.6)
64 (65.3)
0.006
Maternal 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 years , median
(range)
35 (25.5 - 56.1)
32.6 (22.5 - 45.1)
<0.001
Education level mother (high ), n (%)
77 (39.5)
52 (53.1)
0.034
Education level father (highd)a, n (%)
74 (38.7)
53 (56.4)
0.005
Smoking during pregnancy, n (%)
23 (11.8)
5 (5.1)
0.091
5 (2.6)
4 (4.1)
0.489
4 (0 - 5)
0 (0 - 1)
<0.001
Breastfed for > 6 weeksa, n (%)
Parental characteristics
a
d
Alcohol consumption during pregnancy, n (%)
5
Fertility parameters
Time to pregnancy (categoricale)a, median (range)
Fisher’s exact tests, Mann-Whitney U -tests or Student’s t -tests were performed to investigate differences between groups.
Subfertile group: COH-IVF, MNC-IVF and Sub-NC groups together.
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.
a
Birth weight for gestational age is less than -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.
c
University education or vocational colleges.
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.
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: 81
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
5
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
82
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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
83
5
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.
5
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
84
Increased time to pregnancy is associated with less optimal neurological condition in 4-year-old singletons, in vitro fertilization itself is not
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.
5
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.
85
86