Follow-up of cognitive and motor development of 10-year-old singleton children born after ICSI compared with spontaneously conceived children

Transcription

1 Human Reproduction Vol.23, No.1 pp , 2008 Advance Access publication on November 2, 2007 doi: /humrep/dem257 Follow-up of cognitive and motor development of 10-year-old singleton children born after ICSI compared with spontaneously conceived children L. Leunens 1,3, S. Celestin-Westreich 1, M. Bonduelle 2, I. Liebaers 2 and I. Ponjaert-Kristoffersen 1 1 Developmental and Lifespan Psychology, Vrije Universiteit Brussel, 1050 Brussels, Belgium; 2 Academisch Ziekenhuis, Centre for Medical Genetics, Vrije Universiteit Brussel, 1090 Brussels, Belgium 3 Correspondence address. Pleinlaan 2, B-1050 Brussels, Belgium. lize.leunens@vub.ac.be BACKGROUND: This is the first follow-up study of the cognitive, psychosocial, family-relational and medical development of ICSI children at ages 8 and 10. Second-wave study results on the cognitive and motor development of 10-year-old ICSI children are discussed and compared with the outcome at 8 years. METHODS: Developmental outcomes of year-old singletons born through ICSI after at least 32 weeks of gestation were compared with those of 90 singletons of the same age born after spontaneous conception (SC). The outcomes were also compared with the outcomes of the children at age 8. RESULTS: 10-year-old ICSI children obtained total, verbal and performance intelligence scores comparable to those of SC children. No significant differences were found between ICSI and SC children regarding overall motor, manual and ball skills. ICSI children and girls appear to have better balance skills than SC children and boys, respectively. With regard to long-term follow-up, the IQ results of ICSI and SC children at age 10 appear to have converged (from slightly higher scores in the ICSI children at age 8), probably indicating a decreased effect of maternal educational level or stimulating home environment in the ICSI group over time. CON- CLUSIONS: In this follow-up study, ICSI and SC children show a comparable cognitive and motor development until the age of 10. These findings are in line with those obtained at age 8. Keywords: child follow-up; ICSI; cognitive; motor development Introduction Over the last few years, several follow-up studies have been published on the health and development of children born after artificial reproduction techniques (ARTs), and ICSI in particular. These have been mostly prompted by concerns or ambiguous findings on the outcome of these children, making this population the focus of continuing attention. In addition to the studies focusing on the medical outcome, which make up the majority, increasing attention is being given to the developmental/psychological follow-up of these children, yielding mixed to reassuring findings regarding very young to school-aged ICSI children s motor and cognitive development. Overview of previous studies Focusing on the studies that have been carried out on the development of preschool to school-aged children, some large-scale, multi-center studies have been undertaken to investigate the long-term effects of ICSI across diverse medical systems. In this context, no significant differences were found for intelligence quotient (IQ) on the WPPSI-R (Wechsler, 1990) between 5-year-old ICSI and spontaneously conceived (SC) children from Belgium, Sweden and the USA (Ponjaert- Kristoffersen et al., 2004). When analyzing the subtest scores, however, ICSI children appeared more likely to perform better on general factual/acquired knowledge (information) and less well on some visual-spatial abilities (mazes, block design and object assembly), a finding that was attributed to a higher incidence of low birth weight and prematurity in the ICSI study group. A European collaborative follow-up study compared 5-year-old singleton ICSI-born children with IVF and SC controls (Ponjaert-Kristoffersen et al., 2005) recruited in Belgium, Sweden, Denmark, Greece and the UK and matched according to age, sex, birth order, maternal education, parental socioeconomic status and mother s age at birth. Although no significant between-group differences on intelligence came forward, older maternal age at birth was significantly linked to lower full-scale (FSIQ) and verbal IQ (VIQ) in ICSI/IVF children, # The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org 105

2 Leunens et al. as well as lower abilities to anticipate relationships among parts (object assembly) in IVF children. Higher maternal educational level was also linked to better abilities to anticipate relationships among parts (object assembly) in ICSI/IVF children, whereas a low educational level in this group was linked to lower scores on spatial visualization and analysis of wholes into component parts (block design). Overall, these results suggest that factors other than conception mode, especially maternal age at birth and educational level, may prevail in explaining the long-term cognitive development of ICSI children. Continuing follow-up was implemented in our own center, where a proportion of the Belgian populations assessed at age five in the Ponjaert et al. studies (2004, 2005) was reassessed at the age of eight years (Leunens et al., 2006b). This study aimed to provide long-term follow-up and to continue to investigate the relative impact of demographic variables on the development of ICSI children. Total of 151 ICSI and 153 SC children were assessed by means of the Wechsler intelligence scale for children-revised (WISC-R) (Wechsler, 1974; Vander Steene et al., 1986) and the Movement Assessment Battery (ABC) for Children (Henderson and Sugden, 1998), along with an extensive pediatric-neurologic evaluation (Belva et al., 2006) and a family-relational assessment (Leunens et al., 2006b; Leunens et al., submitted). A statistically significant difference was found in intelligence outcome in favor of the ICSI children, although this effect was small and not clinically important and the mean intelligence scores were still situated in similar confidence intervals. Moreover, higher maternal educational level was found to have a positive effect on FSIQ in both groups, which, in combination with a substantial amount of missing data on maternal educational level in the SC group, probably provides an explanation for the higher FSIQ scores in the ICSI group. Additionally, it was hypothesized that the higher educational levels in the ICSI group, combined with these mothers being older when they gave birth, possibly led to a higher degree of child stimulation. It is known that intelligence outcomes are sensitive to stimulation (Neiss and Rowe, 2000). No differences were found between the children s motor abilities. Aim of the current study Taken together, although findings on ICSI children s motor and cognitive development have so far been generally reassuring, further study is needed to clarify inconsistent findings from the existing research literature and to provide follow-up throughout childhood. In response to these needs, this study is the first long-term, two-wave, follow-up investigation into ICSI children s psychological, family-relational and medical development at ages 8 and 10. One of the objectives of this study consisted of the explicit conceptualization of the possible pathways of influence of ICSI to child development. Therefore, different developmental dimensions and socio-cognitive factors of influence were assessed through a multi-method multi-informant approach (Celestin-Westreich, 2004). The present paper discusses the second-wave study findings on the cognitive and motor development of ICSI children at age 10. Overall, given our first-wave study findings on 106 eight-year-old ICSI children, we expected that no effects resulting from the conception mode would be found on ICSI children s cognitive or motor development at ten years. Furthermore, this paper aims to assess the relative impact of demographic variables on children s development. Finally, the present second-wave study findings will be compared with the first-wave results on the intellectual development of ICSI children. Materials and Methods Participants Our study sample (n ¼ 199) consists of Dutch-speaking Belgian children within a larger population that was initially recruited and assessed at the age of eight years and reassessed for this part of the study at the age of 10 years. Singletons born after ICSI (n ¼ 109) and controls born after spontaneous conception (SC) (n ¼ 90) were seen for individual assessment. The children s initial inclusion criteria were: being born after 32 weeks of gestation, singleton and native language Dutch, with at least one European parent. These inclusion criteria were chosen to maximize follow-up rates, to minimize confounding factors due to prematurity and to exclude potential influences due to multiple births, linguistic barriers and socio-cultural differences (see Leunens et al., 2006b). The children were not matched for gestational age, but there was no significant difference in mean gestational age between the study groups (Table I). Gender was fairly evenly distributed, with 56 boys (51.4%) and 53 girls (48.6%) in the ICSI group and 46 boys (51.1%) and 44 girls (48.9%) in the SC group. Procedure The ICSI children were recruited from initial birth cohorts established at the Academic Hospital of the Vrije Universiteit Brussel. They had already been assessed in their second and/or fifth year (Bonduelle et al., 1998; Ponjaert-Kristoffersen et al., 2004, 2005), and they had all taken part in this study s first wave when assessed at age eight, but families who refused participation in the first study wave were not contacted for participation in the second study wave (Leunens et al., 2006b). Close to two-thirds (151/248, 61%) of the eligible Dutch-speaking ICSI cohort turning eight years between February 2001 and December 2003 (n ¼ 248) was tested at eight years. Of the remaining 39% of the cohort (n ¼ 97), 16.5% (41/248) of the families could not be reached at age eight (lost to follow-up), and 22.5% (56/248) refused to participate in this first study wave. Of the 151 ICSI children in our first-wave study population, 109 were reassessed at age 10, resulting in a follow-up rate of 72.2%, or 44% with respect to the initial birth cohort. Of the remaining families, 22 formally refused participation, 12 could not be reached, 6 did not formally refuse but it was impossible to arrange a suitable appointment for them and 2 were unable to come to the hospital due to the mother having a serious disease or a car accident involving the child. At the time of the first study wave, SC control children were selected to match children in the ICSI group (according to age and gender) with further group-level matching for maternal education. The control children were recruited from surrounding schools, with an average response rate of 37.5% at age eight (n ¼ 153). Due to practical reasons, it was possible to recontact 138 of the 153 SC control children from the first study wave for psychological assessment in the follow-up at age 10. Hereof, 90 children were reassessed, resulting in a follow-up rate of 58.8%, whereas 19 families refused further participation, 18 did not respond to the request for participation in the

3 ICSI and development at 10 years Table I. Demographic description of the ICSI and SC study populations. follow-up and 11 were not reached because the child had moved to another school. Demographic variables of children and families who took part in the second study wave ( responders ) were compared with those who did not take part but were seen in the first study wave ( non-responders ), in order to ascertain that the second study wave results were not biased by significant differences between these subpopulations. Both in the ICSI and the SC group, no significant differences were found between responders and non-responders at age 10 for the variables parity, firstborn, gestational age, birth weight, neonatal intensive care unit (NICU) admission, behavioral problems, hospitalization, receiving other therapy, the number of siblings, maternal age at birth, maternal educational level (high, medium and low), paternal educational level (high, medium and low) and marital status (married or cohabiting, new relationship or single). Given the absence of any significant demographic differences between the children and families responding in the second study wave compared with those who did not, it can be concluded that the outcomes reported in this paper are not affected by a participation bias as far as the measured variables are concerned. The full study procedure consisted of the same comprehensive psychological and pediatric-neurological assessment as in the first study wave, covering the child s cognitive and motor development (WISC-R, Movement ABC), the child s perception of its relationship with parents and siblings (FRT), along with a series of physical examinations (including status of prepubertal development, heart auscultation, skin inspection, assessment of weight, height, head circumference, blood pressure, speech, tone, gait, balance, reflexes, strength against resistance, coordination, stereognosis, graphesthesia, hearing, monocular visual acuity and binocular vision). Furthermore, parents were asked to fill out questionnaires concerning their child s ICSI (n ¼ 109) SC (n ¼ 90) P-value Mean SD Mean SD Mean age at testing (months/years) / / P ¼ Firstborn, n, % 91/ /43.3 P ¼ Birth rank P ¼ Only child, n, % 38/34.9 7/7.8 P ¼ Oldest child, n, % 53/ /35.5 P ¼ Middle child, n, % 4/3.7 15/16.7 P ¼ Youngest child, n, % 14/ /40 P ¼ Birthweight (g) P ¼ Gestational age (weeks) P ¼ weeks, n, % 3/2.8 6/ weeks, n, % 106/ /88.9 Unknown, n, % 0 4/4.4 Maternal age at birth (years) P ¼ Unknown, n, % 0 4/4.4 Maternal education level High, n, % 72/ /52.2 P ¼ Medium, n, % 28/ /17.8 P ¼ Low, n, % 1/0.9 3/3.3 P ¼ Unknown, n, % 8/7.3 24/26.6 Paternal education level High, n, % 68/ /43.3 P ¼ Medium, n, % 31/ /24.4 P ¼ Low, n, % 3/2.8 2/2.2 P ¼ Unknown, n, % 7/6.4 27/30 Admission NICU, n, % 23/21.1 3/3.3 P ¼ days,n, % 15/ days,n, % 5/4.6 3/3.3% No of days unknown, n, % 3/2.8 0 Adm status unknown, n, % 6/5.6 26/28.8 behavioral/emotional problems (CBCL), the quality of parental relationship (DAS), their commitment to parenting/to work (Greenberger Scales), parental emotional well-being (GHQ) and perceived parent-child relationship (PARQ). All measures were used with age-appropriate versions where relevant (see Leunens et al., 2006b, for an extensive description). All the children were individually assessed by one single trained psychologist who applied testing conditions, which were similar to the first study wave (e.g. testing times, locations, duration and parental briefing remained unchanged). Although blind testing was not feasible (given that ICSI children were seen in the hospital setting versus school setting for SC children), the use of standardized measures with a fixed study protocol is considered sufficient to minimize potential observer bias (cf. outcome measures). All the participating families provided written informed consent. Outcome measures All the outcome measures were based on standardized tests with satisfactory psychometric properties regarding reliability and validity (Wechsler, 1974; Vander Steene et al., 1986; Henderson and Sugden, 1998). To address the present research questions, cognitive and motor development were measured with the WISC-R and Movement ABC, respectively. The WISC-R The WISC-R (Wechsler, 1974; Vander Steene et al., 1986) is a widely used, individually administered, standardized measure of intelligence for children aged 6 18 years. It yields a FSIQ, VIQ and performance IQ (PIQ) (mean 100, SD 15), along with six verbal scores 107

4 Leunens et al. (information, similarities, arithmetic, vocabulary, comprehension and digit span) and six performance subtest scores (picture completion, picture arrangement, block design, object assembly, coding and mazes) (mean 10, SD 3). Although the Dutch version of the WISC Third Edition (WISC-III) had been published between the first and second study waves in 2002 (Kort et al., 2002), use of the WISC-R was maintained for several reasons. First, taking into account the follow-up study design in which the children had been assessed at age 8 using WISC-R, preference was given to reassessing them with the same instrument in the second study wave for evident reasons of continuity and comparability of the results. This choice was further consolidated by the fact that at the time when data collection of the first study wave started (early 2002), several concerns were raised concerning the construction of the Dutch/Flemish WISC-III norms and the samples used in this regard leading to a temporary retraction of the test followed by later reissuing with revised sample and norm corrections (Celestin-Westreich, 2004). Taken together, study continuation with Dutch language WISC-R was preferred in this context for its demonstrated psychometric qualities. Vander Steene et al. (1986) indeed satisfactorily investigated WISC-R s FSIQ (0.96), VIQ (0.95) and PIQ reliability (0.90), as well as subtest reliabilities (values between 0.87 and 0.68). They also found indications for various forms of validity. The Dutch Commission for Testing matters (COTAN) which issues the most strict evaluations of Dutch-language tests and test versions, judged WISC-R norms as good, reliability as acceptable, face validity as acceptable and criterion validity as insufficient (due to a lack of research) (COTAN, 2002). Finally, the aforementioned arguments were given priority over using the latest test version since the primary aim of this study was to compare cognitive development between two groups being ICSI and SC children, respectively. Furthermore, since intelligence scores cannot be regarded as absolute figures but are always embedded with a certain probability within confidence intervals, the ones used in this paper were those computed by the test constructors of the Flemish/Dutch version of the WISC-R (Vander Steene et al., 1986). Compared with the use of other statistics such as z scores, this method additionally carries the advantage of allowing assessment of the clinical relevance of possible differences in IQ between the groups, which is considered important in this research. The Movement ABC The Movement ABC (Henderson and Sugden, 1998) test measures motor abilities for children aged 5 12 years on three scales: manual skill (e.g. moving pins in a peg board), ball skill (e.g. throwing a beanbag over a 2.5-m distance into a box) and balance (e.g. balancing on each foot on an uneven wooden board for 20 s). The Movement ABC test yields percentile scores with age norms. For Dutch-speaking children at age 10, the 15% percentile cut-off is situated at a total test score of 8.5, with scores below 8.5 indicating above-average motor skills. Statistical analysis Between-factor analyses were carried out on interval or ratio data with two independent variables using ANOVA or t-tests in SPSS 13.0 for Windows. The independent measures used were the child s conception mode (ICSI or SC) and gender. The dependent measures were the WISC-R and Movement ABC test scores. Nominal data were analyzed using Pearson x 2 tests. Significance levels of 0.05 were accepted throughout. In order to compare the first-wave and second-wave study findings with regard to intelligence, the ANOVA general linear method with repeated measures was used, with FSIQ at age 8 versus FSIQ at age 10 as the within-subjects factor, and conception mode as the between-subjects factor. 108 Consistent with our first-wave study, a stepwise linear regression analysis was conducted to control for demographic differences between the conception groups. A detailed outline of the rationale and the steps followed in this analysis can be found in the report of the first-wave outcomes (Leunens et al., 2006b). Results Demographics The children s average age at the time of testing was very similar, but statistical comparison indicated significantly fewer firstborns in the SC group than in the ICSI group (see Table I). Mean birth weight and gestational age were similar in the two groups. Significantly more ICSI children were admitted to a NICU, mostly for short admissions (seven days or less). Thus, the ICSI and control children did not differ in terms of long admissions (over seven days). The mean maternal ages at birth were significantly higher in the ICSI than in the SC group. Maternal educational level, defined as high (postgraduate or a graduate degree), medium (entered university or completely passed school matriculation) and low (partially passed school matriculation or no qualifications at all), did not differ significantly between groups. Cognitive development No significant differences were found between ICSI and SC children s FSIQ, VIQ and PIQ. However, ICSI children obtained lower scores on the similarities and picture completion subtests. Gender did not affect IQ scores in either group. Table II summarizes the IQ scores between the groups and gives detailed subscale scores. In addition and consistent with the first-wave outcomes at age eight, a number of further analyses were carried out to continue our investigation of the effects of demographic data on the long-term development of ICSI children. First, because of (i) the substantial amount of missing data on maternal educational level in the SC group, (ii) the trend toward higher educational levels in the ICSI group and (iii) the documented effects of maternal educational level in previous ICSI research, we repeated the analysis, excluding all children whose maternal educational level was unknown. This analysis resulted in FSIQ outcomes comparable to the initial results (ICSI; SD 12.1; SC SD 12.5; P ¼ 0.443), leading to the assumption that the missing maternal educational level data had no manifest detrimental effects on the initial results. Further, when selectively comparing the FSIQ scores of children with highly educated mothers, no significant differences in FSIQ score between the groups were revealed (ICSI: SD 11.1; SC: SD 12.2; P ¼ 0.521). Second, a stepwise regression analysis was carried out to explore the long-term influence of demographic differences between conception groups on FSIQ. Among those demographic descriptors showing a univariate group difference between the ICSI and SC groups, the following separately entered factors also showed an association with FSIQ: high (standardized b ¼ 0.376, t ¼ 5.198, P ¼ 0.000), medium (standardized b ¼ , t ¼ , P ¼ 0.023) and low (standardized b ¼ , t ¼ ,

5 ICSI and development at 10 years Table II. Mean of the full scale, verbal and performance IQs and subscales as measured by the WISC-R. P ¼ 0.011) maternal educational level. The factors conception mode, maternal age at birth, admission to NICU and firstborn showed no significant association with FSIQ and were therefore excluded from the regression model. Given these results, the regression model (adjusted R 2 ¼ 0.136; P ¼ 0.000) retained the factor high maternal educational level (standardized b ¼ 0.376, t ¼ 5.198, P ¼ 0.000), whereby higher FSIQ is linked to more highly educated mothers. When comparing FSIQ scores at age 8 and 10, an interaction effect between conception mode and FSIQ8 FSIQ10 was noted (F(1, 196) ¼ , P ¼ 0.001). However, as seen from the partial ETA squared measure (0.051), this effect is rather small. In order to define this differential effect of FSIQ8 FSIQ10 in the two conception mode groups, paired samples t-tests were carried out on FSIQ8 FSIQ10 for each conception mode group separately. This was found to be not significant for the SC group (FSIQ , FSIQ , t(1, 89) ¼ 1.152, P ¼ 0.253), but significant for the ICSI group (FSIQ , FSIQ , t(1, 107) ¼ 6.322, P ¼ 0.000). Motor development As shown in Table III, no significant differences are found between ICSI and SC children with regard to total motor, ICSI SC P-value Mean Conf int a SD Mean Conf int a SD Full- scale IQ P ¼ Verbal IQ P ¼ Information P ¼ Similarities P ¼ b Arithmetic P ¼ Vocabulary P ¼ Comprehension P ¼ Digit span P ¼ Performance IQ P ¼ Picture completion P ¼ b Picture arrangement P ¼ Block design P ¼ Object assembly P ¼ Coding P ¼ Mazes P ¼ a Vander Steene et al. (1986); b significant conception mode effect. Table III. Mean total motor score, manual skill score, ball skill score and balance score as measured by the Movement ABC. ICSI SC P-value Mean SD Mean SD Total motor P ¼ score Manual skill P ¼ score Ball skill P ¼ score Balance score P ¼ a a significant conception mode effect; b significant gender effect. P ¼ b handedness or ball skills as measured by the Movement ABC. ICSI children and girls in both groups performed better on balance skills (P, 0.05). Discussion This study aimed to provide insight into the cognitive and motor development of ICSI children followed up to age 10 compared with spontaneously conceived controls. Overall, the results regarding ICSI children s cognitive and motor development remain reassuring for parents and clinicians. Since this is the first follow-up study of ICSI children at the age of 10, these findings cannot be compared with the outcomes of similar studies. Rather, they are compared with the outcomes of the first wave of this study, in which these children were assessed at age eight (Leunens et al., 2006a). Most importantly, 10-year-old ICSI children do not display any significant developmental delays compared with SC children at age 8 or 10, whether it be in terms of cognitive or motor functioning. Therefore, the FSIQ s of ICSI and SC children at 10 years were compared with their first-wave outcomes, revealing two effects. First, the IQ results at age 10 appear to decrease somewhat compared with the results at age 8, meaning that the slightly yet statistically significantly different high mean IQs that were found at age 8 appear to normalize by age 10. After verification of interaction effects, this effect of a relative IQ decrease toward the mean reveals a significant difference only in the ICSI group. Also, or as a result hereof, ICSI and SC children s intelligence scores tend to converge compared with their intelligence scores in the first study wave at age eight. Several considerations may contribute toward interpreting this effect. First, in any case, it should be reminded that children s intelligence scores are situated in the normal range. Furthermore, given that a slight but rather intriguing significant difference appeared between full intelligence outcomes in favor of the ICSI group at age eight, the difference revealed at age eight was not regarded as clinically relevant, as it was attributed to higher maternal educational 109

6 Leunens et al. levels in the ICSI group, in combination with older maternal age. Although this difference appears to wear off at age 10, it should be noted that higher maternal educational level still continues to contribute to higher FSIQ scores. As for the possible role of other influencing demographic variables, those which differed between the two conception groups (firstborn, maternal age at birth and NICU admission) did not appear to play a role in the intellectual development of our study groups at age 10. Moreover, in an effort to control for a possible bias that could be caused by attrition of study participants over the course of the study, the full set of demographic characteristics of children and families who were retained in the second study wave were compared with those who were not (families who refused to participate or were lost to follow-up ). Both in the ICSI and the SC groups, no significant differences could be detected between characteristics of the responders and the nonresponders. The conclusion that intelligence outcomes at age 10, and the difference in IQ scores between age 8 and 10 in the ICSI group, can not be attributed to a bias between responders and non-responders at age 10 caused by this set of demographic factors. Although our results are very reassuring with regard to the development of ICSI-conceived children up to middle elementary school age, generalization of our findings deserves caution given some study limitations pertaining to initial response rates, participation bias and inclusion criteria. One factor that may limit generalizations from these findings is the response rate. Indeed, although a follow-up rate of 72% was achieved with respect to the first study wave, more than half of the initial birth cohort could not be seen in this second study wave, which requires one to question the potential participation bias. This difficulty is evidently inherent to the nature of long-term follow-up studies and underlines the complexity of longitudinal research. Concerning the main preoccupation in this context, namely that dropout rates would contribute to missing out on information for non-responding ICSI children who do present difficulties (in contrast to the participating ICSI children), several indicators appear reassuring to this regard. First, through the analysis of answers to a telephone checklist, no major problems were identified in the first-wave dropout ICSI group (see Belva et al., 2006), Second, as discussed earlier, demographic characteristics of the second study wave dropout ICSI group were compared with those of the families who chose to continue participating in the second study wave, indicating no differences between those two groups and thus suggesting that our second-wave study findings are not biased by participant versus dropout ICSI populations. Moreover, a retrospective analysis was carried out on FSIQ scores at age 8 of the participating versus nonparticipating children at age 10. No significant difference was found in this analysis (FSIQ8(participant age 10) ; FSIQ8(non-participant age 10) ; P ¼ 0.223). Likewise, a participation bias could not be ruled out for the control group given the low follow-up rate of 48.8%, which tends to be typical when participation is voluntary [e.g. recruitment not based on established (medical) structures such as a national birth registry existing in other countries, no financial rewards]. 110 The SC group may have felt less concerned by the research and therefore probably less compelled to continue participation in the second study wave. As a result, the recruited control group may have been biased toward children with either fewer or more problems compared with non-responders. Therefore, the same retrospective analysis was carried out on FSIQ scores at age 8 of the participating versus non-participating SC children at age 10. This indicated that non-participant controls at age 10 obtained significantly lower FSIQ scores at age 8 than controls who continued to participate at age 10 (FSIQ8 (participant age 10) ; FSIQ8(non-participant age 10) ; P ¼ 0.001). Therefore, it can be hypothesized that the recruited SC control group in the second study wave would have been biased toward children with few problems, which could be an explanation for the fact that the decrease in FSIQ with age for the SC children s FSIQ was not significant. Our findings on ICSI and SC children s cognitive and motor development are also based on the use of statistically normed instruments, thus allowing comparison with normative data for Dutch-speaking Belgian children. Comparison of the mean and SD IQ s in our sample with the Belgian norms indicated that scores of the participating ICSI children are situated in the (above) average range. The generalization of our findings may also be limited given that only children born after 32 weeks of gestation were included. It must be noted that this inclusion criterion still contains children who were born as premature infants between 32 and 37 weeks of gestation, who s IQs were average to high in both groups (ICSI ; SC , P ¼ 0.929) regardless of possible major birth defects and low birth weights, which were not excluded per se from this study. Since this is the first study specifically designed to ensure long-term follow-up of the effects of the ICSI procedure on child development, these results are new and certainly call for replication with later cohorts, including those with more socio-culturally diverse populations, in other countries, and across other medical systems. Given the pilot nature of research on ICSI youth, continuing follow-up is undoubtedly recommended to make sure they continue to do well. For instance, in addition to the ultimate question of these children s own reproductive future, as ICSI youth reach puberty and adolescence, a comprehensive perspective on their development calls for increasing attention to their personality development and their own perception of growing up as part of an ART family, and how this potentially influences their family relationships. Acknowledgements We would like to thank all the families and the schools who kindly agreed to take part in the study. Funding This study was supported by a research grant from the Fonds voor Wetenschappelijk Onderzoek Vlaanderen (FWO) and the Onderzoeksraad Vrije Universiteit Brussel.

7 ICSI and development at 10 years References Belva F, Henriet S, Liebaers I, Van Steirteghem A, Celestin-Westreich S, Bonduelle M. Medical outcome of 8-year-old singleton ICSI children (born. ¼ 32 weeks gestation) and a spontaneously conceived comparison group. Hum Reprod 2007;22: Bonduelle M, Joris H, Hofmans K, Liebaers I, Van Steirteghem A. Mental development of 201 ICSI children at 2 years of age. Lancet 1998;351:1553. Celestin-Westreich S. Behavioral and Family Assessment in Children. Leuven: Acco, COTAN. Documentatie an tests en testresearch in Nederland. Amsterdam: Boom test uitgevers, Henderson SE, Sugden DA. Movement Assessment Battery for Children. Handleiding. Lisse, The Netherlands: Swets & Zeitlinger B.V, Kort W, Compaan EL, Bleichrodt N, Resing WCM, Schittekatte M, Bosmans M, Vermeir G, Verhaeghe P. WISC-III NL. Handleiding. London: The Psychological Corporation, 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 2006a;21: Leunens L, Celestin-Westreich S, Bonduelle M, Ponjaert-Kristoffersen I. Influence of ICSI on parenting stress and child behavior at age 8. Hum Reprod 2006b;21(Suppl 1):113. Neiss M, Rowe DC. Parental education and child s verbal iq in adoptive and biological families in the national longitudinal study of adolescent health. Behav Genet 2000;30: Ponjaert-Kristoffersen I, Tjus T, Nekkebroeck J, Squires J, Verté D, Heimann M, Bonduelle M, Palermo G, Wennerholm UB, Collaborative study of Brussels, Gotheborg New York. Psychological follow-up study of 5-year-old ICSI children. Hum Reprod 2004;19: Ponjaert-Kristoffersen I, Bonduelle M, Barnes J, Nekkebroeck J, Loft A, Wennerholm UB, Tarlatzis BC, Peters C, Hagberg BS, Madsen SV et al. An International collaborative study of ICSI conceived 5 year olds, child outcomes: cognitive and motor assessment. Pediatrrics 2005;115:E283 E289. Vander Steene G, van Haasen PP, De Bruyn EEJ, Coetsier P, Pijl YL, Poortinga YH, Lutje Spelberg HC, Spoelders-Claes R, Stinissen J. Wechsler Intelligence Scale for Children-Revised. Nederlandstalige Uitgave. Lisse: Swets & Zeitlinger B.V, Wechsler D. Wechsler Intelligence Scale for Children-Revised: Manual. New York: The Psychological Corporation, Wechsler D. Manual for the Wechsler Preschool and Primary Scales of Intelligence-Revised. New York: The Psychological Corporation, Submitted on January 16, 2007; resubmitted on May 24, 2007; accepted on May 31,