Neonatal outcome in a Danish national cohort of 3438 IVF/ICSI and 10 362 non-IVF/ICSI twins born between 1995 and 2000

Anja Pinborg1,2, Anne Loft1, Steen Rasmussen3, Lone Schmidt4, Jens Langhoff-Roos5, Gorm Greisen6 and Anders Nyboe Andersen1

1 The Fertility Clinic, 5 Department of Obstetrics and Gynecology, 6 Department of Neonatology, University of Copenhagen, Rigshospitalet, 3 National Board of Health, Health Statistics, Copenhagen and 4 Institute of Public Health, University of Copenhagen, Denmark

2 To whom correspondence should be addressed at: The Fertility Clinic, University of Copenhagen, Rigshopitalet, Blegdamsvej 9, DK-2100 Copenhagen Ø, Denmark. e-mail: apinborg{at}rh.dk


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: In Denmark, one-third of twin pregnancies are the result of IVF/ICSI treatment. Limited data on neonatal outcome in IVF/ICSI twins are available in the literature. METHODS: A register study was conducted on neonatal morbidity and mortality in a complete national twin cohort including all 3438 (3393 live-born) IVF/ICSI and 10 362 (10 239 live-born) non-IVF/ICSI twins born between 1995 and 2000. Twins were identified in the National Medical Birth Registry and dichotomized into IVF/ICSI and non-IVF/ICSI by cross-reference with the Danish IVF Registry. Data on neonatal morbidity and mortality were retrieved from the Danish Patient Registry and the Danish Registry of Causes of Deaths. In order to exclude monozygotic twins, sub-analyses on unlike-sex twins were conducted. RESULTS: A birth weight discordance of >20% was observed in 20.6% of IVF/ICSI versus 15.7% of control twin pairs (P < 0.001). The risk of discordant birth weight >20% was OR 1.29 (95% CI 1.04–1.58) in unlike-sex IVF/ICSI twins versus control twins. The risk of delivery at <37 completed weeks and birth weight <2500 g was similar in the two cohorts; however, in unlike-sex IVF/ICSI versus control twins the risk of delivery at <37 weeks and birth weight <2500 g was OR 1.22 (95% CI 1.09–1.38) and OR 1.25 (1.11–1.40) respectively. After stratification for maternal age and parity, these risks disappeared. IVF/ICSI twins carried a higher risk of admittance to a neonatal intensive care unit (NICU) than control twins (OR 1.18, 95% CI 1.09–1.27), and this was even more pronounced in unlike-sex twins [OR 1.34 (95% CI 1.19–1.51)]. No differences were observed in malformation or mortality rates between the two cohorts. CONCLUSIONS: Despite higher birth weight discordance and more NICU admissions among IVF/ICSI twins, neonatal outcome in IVF/ICSI twins seems to be comparable with that of non-IVF/ICSI twins, when only dizygotic twins were considered in the comparisons.

Key words: clinical epidemiology/congenital malformations/ICSI twins/IVF/neonatal outcome


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In some European countries, up to 4% of all infants are born after assisted reproductive technology (ART), implying that overall national perinatal statistics are affected by the contribution of ART (Nygren and Nyboe Andersen, 2002Go). Although a relatively small proportion of the population uses ART, the introduction of these techniques has imposed worldwide increases in multiple birth rates. Between 1970 and 1996, a 2-fold increase in the twin birth rate was observed in Denmark (Imaizumi, 1998Go). The annual ESHRE report from 1999 indicated that 24% of IVF and ICSI deliveries in Europe were twin deliveries, and consequently almost 40% of all IVF and ICSI babies are born as twins (Nygren and Nyboe Andersen, 2002Go).

Twin pregnancies are the main factor in the overall poorer perinatal outcome in IVF pregnancies, even though surveys on IVF singleton pregnancies have found slightly higher complication rates compared with spontaneously conceived singleton pregnancies (Bergh et al.,1999; Dhont et al., 1999Go; Westergaard et al., 1999Go; Schieve et al., 2002Go). A Danish register study with 854 IVF twins observed significantly higher birth weight in IVF twins than in spontaneously conceived twins (Westergaard et al., 1999Go), whereas other surveys have reported similar perinatal morbidity and mortality in IVF/ICSI and spontaneously conceived twins (Olivennes et al., 1996Go; Bernasko et al., 1997Go; Fitzsimmons et al., 1998Go; Dhont et al., 1999Go; Koudstaal et al., 2000Go). In 1999, one group (Dhont et al., 1999Go) observed no differences in perinatal outcome between 1241 dizygotic (DZ) IVF and naturally conceived twin pregnancies. However, a later survey on 480 DZ IVF twin pregnancies observed poorer perinatal outcome in IVF compared with naturally conceived twin pregnancies (Lambalk and van Hooff, 2001Go). The frequency of monozygotic (MZ) babies in IVF twins is 1–2% compared with 30% in spontaneously conceived twins (Schachter et al., 2001Go; Sperling and Tabor, 2001Go). Previous studies on spontaneously conceived twins have found perinatal morbidity and mortality to be higher in MZ than in DZ twins (Bryan et al., 1987Go; Loos et al., 1998Go). Thus, it could be assumed that IVF twin pregnancies carry less obstetric risks than spontaneously conceived twin pregnancies.

In severe but rare events such as malformations and perinatal death, a large study group is needed to investigate the risk with sufficient power. The aim of the present study was to assess the outcome of a complete Danish IVF/ICSI twin birth cohort from 1995 to 2000 compared with all non-IVF/ICSI twin births in Denmark during the same period. In addition, sub-analyses on unlike-sex twins were performed to exclude all MZ—and thus monochorionic twins—from both cohorts. Data from the Danish IVF registry were cross-linked with existing registries on births, malformations, admissions to the neonatal intensive care unit (NICU) and deaths.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This population-based cohort study included all twins born in Denmark after IVF or ICSI between January 1, 1995 and December 31, 2000. The control group consisted of all non-IVF/ICSI twins born in Denmark during the same 6-year period.

The National Medical Birth Registry (MBR) recording all live births and stillbirths in Denmark was used to identify all women, who gave birth to live-born and stillborn twins between January 1, 1995 and December 31, 2000 (Knudsen et al., 1998). In Denmark, every individual has a unique identification number in the Centralised Civil Registry (CPR number) that is widely used by healthcare and social-service providers. The CPR number in the MBR identified each woman. Subsequently, those women who delivered after IVF or ICSI treatment were identified by cross-reference with data recorded in the Danish IVF Registry. This enabled the twin mothers to be dichotomized into the study population, which included women with IVF/ICSI twins and a control group of women with non-IVF/ICSI twins.

In Denmark, since January 1, 1994 it has been compulsory to register each initiated IVF or ICSI cycle with the Danish IVF Registry in the National Board of Health (Nyboe Andersen et al., 1999Go). Records on infertility history and fertilization method were retrieved from the Danish IVF Registry. The CPR number of each individual infant in the two cohorts was identified through a unique existing linkage between the CPR number of a mother and her children in the MBR. Information on previous deliveries, duration of pregnancy in completed weeks, birth weight (in grams), rate of Caesarean section and survival of babies was retrieved from the MBR. By using record linkage with the Danish Patient Registry and the Danish Cancer Registry, admissions to the NICU, congenital malformations and cancers were identified (Andersen et al., 1999Go). Mortality rates were retrieved from the National Registry of Causes of Deaths, while diagnoses of cancer were retrieved from the Danish Patient Registry and cross-checked in the Danish Cancer Registry. Diagnoses including cancers were classified according to the International Classification of Diseases, 10th Edition (ICD-10) in the Danish Patient Registry. The quality of data on congenital abnormalities from the Danish Patient Registry has been assessed in a recent study and found to be acceptable for general surveillance and epidemiological research (Larsen et al., 2003Go). In the MBR, the duration of gestation in IVF/ICSI twin pregnancies is determined from 14 days before the date of oocyte aspiration, and in control twin pregnancies from the first day of the last menstrual period.

A total of 3438 IVF/ICSI twins (3393 live-born, 45 stillborn) and 10 362 control twins (10 239 live-born, 123 stillborn) was identified. Stillborn babies were excluded from further statistical analyses regarding neonatal outcome and congenital malformations, as the register recording of the stillborn was inadequate. Only CPR numbers were used in all linkages. The study was approved by the Danish Data Protection Agency and assessed by the Scientific Ethical Committee of Copenhagen and Frederiksberg Municipalities, who recorded no objections.

Definitions
‘Delivery’ was defined as live-born and stillborn babies delivered after 22 completed weeks of gestation, while ‘duration of gestation’ in IVF/ICSI twin pregnancies was defined from 14 days before the date of egg aspiration, and in control twin pregnancies from the first day in the last menstrual period. A pre-term birth was defined as delivery before 37 completed weeks of gestation. Low birth weight was defined as <2500 g, and very low birth weight as <1500 g.

‘Neonatal mortality’ was defined as the number of children who died <28 days after delivery per 1000 children born with a gestational age of 22 weeks or more. Intrauterine and intrapartum deaths were not included. ‘Infant mortality’ was defined as live born children who died within the first year.

‘Malformations’ were defined as conditions registered in the International Classification of Diseases and Health Related Problems, 10h Revision (ICD-10; Danish National Board of Health, 1993), as a congenital malformation or chromosome abnormality (ICD-10: Q00-Q99). ‘Major congenital malformations’ were defined as malformations that generally cause functional impairment or required surgical correction. The remaining malformations were considered ‘minor’.

‘Birth weight discordance’ was defined as the weight difference between the smaller and larger child of a pair of twins, expressed as the percentage of the birth weight of the larger child.

Statistical analysis
The results were analysed using SPSS (Statistical Packages for Social Sciences) version 10.0. Statistical significance was defined as P < 0.05. Differences in the means of continuous parametric data were analysed using Student’s t-test. {chi}2 analyses were used to compare distributions between the two groups.

With regard to analyses on gestational age, ‘twin pregnancy’ was the unit of analysis, whereas in terms of birth weight and malformations, ‘infant’ was the unit of analysis. Birth weight was calculated as the mean of two twins in a pair, and malformations were registered for both twins in a pair. Twin data for analysis was either entered as twin individuals, or as a twin delivery. The statistical problem caused by ties between a twin pair (e.g. similar birth weights, similar malformations, etc.) was always present and unavoidable, but was always taken into consideration.

Odds ratios were calculated using the Mantel–Haenszel estimate after stratification. For stratification of data, the following maternal characteristics were used: age (5-year age-groups of <30, 30–34, 35–39 and ≥40 years) and parity (0 or ≥1 previous deliveries). In terms of malformations, data were stratified for treatment method (IVF versus ICSI). Data on admissions to the NICU were stratified for duration of pregnancy in completed weeks (<32, 32–36 and >37 weeks).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Demographic data
The maternal characteristics of the women are listed in Table I. The age and parity distribution of women who delivered IVF/ICSI twins differed strikingly from that of women in the control twin group; as expected, IVF/ICSI twin mothers were ‘elderly’ primiparae. Of the IVF/ICSI twin mothers, 75.4% had received IVF and 24.6% ICSI treatment.


View this table:
[in this window]
[in a new window]
 
Table I. Demographic data on women and children in the three cohorts
 
The frequency of boys was similar among IVF/ICSI twins (52.1%) and control twins (52.6%). Of the IVF/ICSI twins, 50.8% were same-sex twin pairs, in contrast to 65.4% same-sex control twin pairs (P < 0.001). The zygosity of twins was determined using Weinberg’s differential method (Weinberg, 1902Go). Doubling the number of opposite-sex twins gave an estimate of the total number of DZ twins, the remainder being MZ. This calculation assumed that the sex ratio of DZ twins was 50:50. In the present study, 825 opposite-sex IVF/ICSI twin pairs out of 1676 twin pairs were observed for which the sex of both twins was known. This gave an estimate of 26 (1.6%) MZ twin pairs in the IVF/ICSI twin group. The corresponding figures in the control twin group were 1773 opposite-sex twin pairs out of 5103 twin pairs, which gave an estimate of 1557 (31%) MZ twin pairs in the control twin group.

Birth weight and gestational age
Gestational age and birth weight for children born after 22 completed weeks are presented in Table II. Mean birth weight and gestational age was significantly lower for IVF/ICSI twins compared with control twins. The unadjusted risk of birth weight <2500 g was similar, but the risk of delivery at <37 weeks was significantly higher in IVF/ICSI twins compared with controls (Table III). After stratification for maternal age, the risk of low birth weight and preterm birth became significantly higher in IVF/ICSI twins compared with controls. In subsequent analyses with stratification for both maternal age and parity, the differences disappeared and IVF/ICSI twins even had a lower risk of birth weight <2500 g compared with control twins [Odds Ratio (OR) 0.88, 95% CI 0.80–0.96)] (Table III).


View this table:
[in this window]
[in a new window]
 
Table II. Gestational age and birth weight in live-born children with gestational age of 22 completed weeks or more
 

View this table:
[in this window]
[in a new window]
 
Table III. Odds ratios (OR, 95% CI) for low birth weight and short duration of gestation in IVF/ICSI twins (n = 3393) versus control twins (n = 10 239) and in opposite-sex IVF/ICSI twins (n = 1650) versus opposite-sex control twins (n = 3546)
 
In order to exclude MZ twins, sub-analyses on unlike-sex twin pairs among IVF/ICSI twins (n = 1650) and control twins (n = 3546) were performed, and results indicated that mean birth weight and gestational age were significantly lower in unlike-sex IVF/ICSI twins than in unlike-sex control twins. As observed for the total twin groups, the discrepancies between the unlike-sex twin groups remained after stratification for maternal age only, but disappeared after stratification for both maternal age and parity (Table III).

Birth weight discordance
The mean birth weight discordance was 12.5% in IVF/ICSI twins and 11.3% in control twins (P < 0.001). The risk of discordant birth weight in unlike-sex and same-sex IVF/ICSI twins was equal, whereas the risk of discordant birth weight in unlike-sex control twin pairs was higher than in same-sex control twin pairs (OR 1.25, 95% CI 1.07–1.47). A birth weight discordance of more than 20% was observed in 20.6% of the IVF/ICSI twin pairs, and in 15.7% of the control twin pairs (P <0.001). Unlike-sex IVF/ICSI twins were more likely to have discordance >20% than unlike-sex control twins (OR 1.29, 95% CI 1.04–1.58).

Congenital malformations
The number of major and minor birth defects in different organ systems is presented in Table IV. A complete list of malformations with exact diagnoses is available from the first author. Data on malformations were collected from date of birth until December 31, 2001; thus, the minimum period of follow-up was one year. Among the 3393 live-born IVF/ICSI twins, 211 (62.1/1000) children with 250 congenital malformations were recorded. Of the 211 IVF/ICSI twin infants with congenital malformations, 164 of 2530 were conceived after IVF and 47 of 824 after ICSI treatment. No significant difference was observed in the total malformation rates between IVF and ICSI infants. Four specific conditions were examined which were earlier reported to be associated with a higher incidence in ART children, namely neural tube defects, alimentary atresia, omphalocele with IVF-conceived children, and hypospadias with infants conceived after ICSI. Three cases of neural tube defects were recorded (all three in IVF, none in ICSI children), seven alimentary atresia (seven in IVF, none in ICSI children), no IVF/ICSI infants with omphalocele, and finally nine hypospadias (five in IVF, four in ICSI children) were identified in the IVF/ICSI cohort. Statistical comparisons of these specific conditions revealed no increased incidence in IVF/ICSI twins versus controls; neither was the incidence of hypospadias higher in ICSI versus IVF infants.


View this table:
[in this window]
[in a new window]
 
Table IV. Number of major and minor congenital malformations and rates per 1000 children in live-born children with gestational age ≥22 weeks
 
Among the 10 239 live-born control twins, 596 (58.2/1000) children with 755 malformations were identified (Table IV). There was no increased risk of malformations in IVF/ICSI versus control twins, neither between the total cohorts (OR 1.08, 95% CI 0.92–1.28) nor in the sub-analyses on unlike-sex twins (OR 1.24, 95% CI 0.97–1.58).

Mode of delivery
The Caesarean section (CS) rate was 52.9% in IVF/ICSI twin and 42.7% in control twin deliveries; moreover, the risk of CS in IVF/ICSI twin deliveries was 1.5-fold higher than that in control twin deliveries (OR 1.5, 95% CI 1.3–1.7). This risk disappeared after adjustment for maternal age and parity (OR 1.1 95% CI 1.0–1.2).

NICU admissions
Among the IVF/ICSI twins and control twins, 56.3% and 52.4% respectively of the children were admitted to the NICU. IVF/ICSI twins carried a higher risk of admittance to NICU than control twins (OR 1.18, 95% CI 1.09–1.27). Stratification for gestational age yielded the same increased risk for IVF/ICSI versus control twins (OR 1.17, 95% CI 1.06–1.28). By restricting data to only DZ twins, sub-analyses on unlike-sex twins were carried out. The risk of admittance to NICU was even higher for unlike-sex IVF/ICSI compared with unlike-sex control twins (OR 1.34, 95% CI 1.19–1.51), and this higher risk was maintained after adjustment for gestational age (OR 1.30, 95% CI 1.13–1.50).

The average number of days spent in the NICU was 19.8 and 18.4 (P = 0.003) for IVF/ICSI twins and control twins respectively. The frequency of children hospitalized for more than 7 days was 74.6% and 69.9% (P <0.001), and for more than 28 days was 27.6% and 24.6% (P = 0.02) for IVF/ICSI twins and control twins respectively.

Mortality
Mortality rates in the IVF/ICSI twin group and the two control groups are presented in Table V. No significant difference in stillborn rate (P = 0.6) and neonatal mortality rate (P = 0.3) was seen between the IVF/ICSI and control twins. The total infant death rate (live-born children who died within the first year of life) was significantly lower among IVF/ICSI twins (10/1000) compared with control twins (15/1000) (P = 0.04). For unlike-sex twins, this mortality rate was 12/1000 in the IVF/ICSI twin group and 11/1000 in the control twin group (P = 0.65).


View this table:
[in this window]
[in a new window]
 
Table V. Mortality rates among IVF/ICSI and control twins
 
Cancer incidence
Among the 3393 twins born after IVF, none developed cancer, while among 10 239 control twins 11 children (1.1/1000 children; P = 0.076) developed cancer before December 31, 2001. The minimum period of observation was one year. The cancer diagnoses were acute lymphatic leukaemia, astrocytoma, and non-specific tumours in the heart, cerebrum, suprarenal gland and soft tissues.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
To the best of the present authors’ knowledge, this is the largest controlled survey on neonatal outcome in IVF/ICSI twins. It is a population-based cohort study consisting of all twins born after IVF or ICSI treatment in Denmark from 1995 to 2000. Moreover, all spontaneously conceived twins born during the same 6-year period were included as the control group. This national survey has several advantages. First, the compulsory Danish IVF Registry and unique CPR number system enabled all IVF and ICSI twin births from 1995 to 2000 to be traced. Second, the linkage between women and children in the National Medical Birth Registry (MBR) allowed the identification of all individual infants in the three cohorts, and for information to be retrieved from other registries such as the Danish Patient Registry, the National Registry on Causes of Deaths, and the Danish Cancer Registry. Moreover, the confounding effects of maternal age and parity were minimized by stratification, which also made it possible to investigate the role of preterm labour on neonatal morbidity between the two cohorts. As data on zygosity were unavailable in the MBR, separate analyses were performed on unlike-sex twin pairs to exclude MZ twins.

The limitations of the present study were the lack of information on the use of ovarian stimulation among spontaneously conceived twins, and also on socioeconomic status. In order to compensate for this restriction in the register data, a questionnaire survey was carried out which included all those women in the three cohorts who gave birth in 1997. Data from this study showed no differences in socioeconomic status between the women in the three different groups (Pinborg et al., 2003bGo). Moreover, the results indicated that 17.3% of women in the control twin group had a history of ovarian stimulation, with or without intrauterine insemination (IUI) prior to delivery. Anticipating that the results from 1997 could be extrapolated to the entire period from 1995 to 2000, an average of 15–20% women conceived after ovarian stimulation in the ‘spontaneously’ conceived twin group in the present study (Pinborg et al., 2003aGo). Although recent studies have not reported any influence of the stimulation procedure on neonatal outcome in either singletons or twins (Olivennes et al., 1996Go; Daniel et al., 2000Go; Källen et al., 2002Go), the lack of identification of those women who underwent ovarian stimulation was a methodological problem not only in the present study but also in the vast majority of other register studies.

The rate of MZ twin pregnancies is lower in IVF twin pregnancies, namely 1–2% versus 30% in spontaneously conceived pregnancies (Schachter et al., 2001Go; Sperling and Tabor, 2001Go). It is known that morbidity and mortality is higher in MZ twins and mainly in monochorionic twins (Bryan et al., 1987Go; Loos et al., 1998Go). As expected, in the present study the estimated proportion of MZ twins in IVF/ICSI pregnancies was lower (1.6%) than in control pregnancies (31%).

In contrast to the assumption of fewer obstetric risks in IVF/ICSI twins due to a lower MZ rate, comparable risks of low birth weight and gestational age in the total cohorts of IVF/ICSI and control twins were observed. Only the risk of birth weight <2500 g was significantly lower in IVF/ICSI twins after age and parity stratification. In unlike-sex twins, the risk of low birth weight and preterm birth was significantly higher in IVF/ICSI twins than in control twins but, as expected, these risks disappeared after correction for maternal age and parity. With regard to the analyses of gestational age, it should be noted that this variable in the MBR is registered in units of completed weeks and that the exact gestational age in days is not provided. Thus, one point of discussion is whether or not gestational age is a sufficiently precise parameter on which to base the conclusion that there is no difference in gestational age between IVF/ICSI and control twin pregnancies.

It was of interest that stratification for maternal age in both analyses on the total twin and the unlike-sex twin cohorts seemed to increase the risk in the IVF/ICSI twin group. After stratification for both maternal age and parity, the risk in the IVF/ICSI twin group decreased; this finding was in accordance with a previous report that maternal age was the least influencing factor on outcome in ART twins, and that parity was the strongest independent predictor of outcome (Daniel et al., 2000Go).

Although a similar obstetric outcome was observed in IVF/ICSI and spontaneously conceived twins, the former were more likely to be admitted to the NICU. Even after stratification for gestational age, the unlike-sex IVF/ICSI twins had a 1.3-fold higher risk of NICU admittance. Although this might reflect the fact that healthcare professionals take more precautions when treating twins born after IVF or ICSI treatment, an increased neonatal morbidity cannot be excluded. In accordance with more precautions being taken in women who conceived after ART, or in their children, the results of the previous questionnaire survey indicated that IVF/ICSI twin mothers more often required sick leave and hospitalization during pregnancy than control twin mothers, though their morbidity during pregnancy was not higher (Pinborg et al., 2003cGo).

In line with the results of previous studies on birth weight discordance, an increased risk of discordance >20% was observed in IVF/ICSI versus control twins (Bernasko et al., 1997Go; Daniel et al., 2000Go; Koudstaal et al., 2000Go). Previously, it has been shown that birth weight discordance is higher in spontaneously conceived MZ compared with DZ twin pregnancies (Rydhstroem, 1996Go), but this has been explained by a net imbalance in directional blood flow through placental anastomotic vessels (Bernasko et al., 1997Go). Hence, it is difficult to interpret the present finding of a higher risk of discordance in unlike-sex versus same-sex control twins, but this was also observed by others (Bernasko et al., 1997Go). One possibility is that same-sex control twin pairs consist of both MZ and DZ twins, and a plausible explanation for the difference in fetal growth in DZ pregnancies is genetic dissimilarity.

Previous reports on neonatal mortality in IVF/ICSI versus spontaneously conceived twins have yielded contradictory results (Bergh et al., 1999Go; Dhont et al., 1999Go; Westergaard et al., 1999Go; Koudstaal et al., 2000Go; Lambalk and van Hooff, 2001Go; Koivurova et al., 2002Go). Similar to the majority of studies, the present results showed no significant differences in stillbirth and total mortality rates between IVF/ICSI and control twins. However, a significantly higher mortality rate was found during the first year after delivery in the total control twin cohort than in IVF/ICSI twins (P = 0.04). The majority of control twins died within the first week of life, when the biggest difference in mortality rates between IVF/ICSI and control twins was seen (Table V). This higher mortality rate of control twins within the first year of life was not present in the comparison restricted to unlike-sex IVF/ICSI and control twins, presumably while the MZ control twins were excluded.

No comparative reports exist on increased malformation rates in IVF/ICSI twins and spontaneously conceived twins. In the present study, the malformation rate was similar in both total cohorts and also when the data were restricted to unlike-sex twins.

Patent ductus arteriosus (PDA) is a developmental anomaly which has strong associations with preterm birth. As in a previous Swedish register study, the decision was made here to define PDA as a minor congenital malformation (Ericson and Källén, 2001Go). Excluding this anomaly from the analyses implied no changes in the differences in malformation rates between IVF/ICSI and control twins in the present study.

Separate analyses on the infants in IVF/ICSI twin cohort produced similar total malformation rates in infants conceived after IVF and ICSI treatment, and this finding was in agreement with a previously published report (Bonduelle et al., 2002Go). The Swedish register study yielded significant differences in the incidence of specific conditions (neural tube defects, hypospadias, alimentary tract atresia, omphalocele) between IVF and ICSI and spontaneously conceived children (Ericson and Källén, 2001Go). In the light of these findings, these specific conditions were examined in detail and no differences were observed either between IVF/ICSI infants and controls or between IVF- and ICSI-conceived children. It should be taken into account that the low number of children with specific malformations provides the present study with a very low power to detect such differences. Except for congenital malformations, separate analyses were not performed for IVF and ICSI children, as no previous studies have assessed differences in obstetric outcome between IVF and ICSI deliveries (Bergh et al., 1999Go; Bonduelle et al., 2002Go).

Consistent with a previous survey on cancer incidence in IVF children, no difference was observed between IVF/ICSI and control twins, though any statistical comparison should be carefully considered as only very few cases were observed (Bruinsma et al., 2000Go).

In conclusion, the results of the present study suggest that neonatal outcome in twins conceived after IVF or ICSI treatment is generally comparable with that in twins after spontaneous conception. However, IVF/ICSI twins were more likely to be of discordant birth weight and to be admitted to the NICU. Even in unlike-sex twins stratified for maternal age and parity, these were the only differences found. Hence, the present findings indicate that IVF/ICSI twins do not carry the slightly increased risk of adverse outcome seen in IVF/ICSI singletons compared with spontaneously conceived singletons, even after maternal age and parity has been taken into account.


    Acknowledgements
 
The authors thank Lone Mortensen, Health Statistics, Danish National Board of Health for help in providing data from the Danish IVF Registry. The study was supported financially by grants from the Danish Medical Research Council, the Danish Hospital Foundation for Medical Research, Region of Copenhagen, The Faroe Islands and Greenland, and the Research Foundation of Queen Louise’s Paediatric Hospital.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Andersen TF, Madsen M, Jorgensen J, Mellemkjaer L and Olsen J (1999) The Danish Hospital Register. A valuable source of data for modern health sciences. Dan Med Bull 46,263–268.[ISI][Medline]

Bergh T, Ericson A, Hillensjö T, Nygren K-G and Wennerholm U-B (1999) Deliveries and children born after in-vitro fertilisation in Sweden 1982–95: a retrospective cohort study. Lancet 354,1579–1585.[CrossRef][ISI][Medline]

Bernasko J, Lynch L, Lapinski R and Berkowitz L (1997) Twin pregnancies conceived by assisted reproductive techniques: maternal and neonatal outcomes. Obstet Gynecol 89,368–372.[Abstract/Free Full Text]

Bonduelle M, Liebaers I, Deketelaere V, Derde M-P, Camus M, Devroey P and Van Steirteghem A (2002) Neonatal data on a cohort of 2889 infants born after ICSI (1991–1999) and of 2995 infants born after IVF (1983–1999). Hum Reprod 17,671–694.[Abstract/Free Full Text]

Bruinsma F, Venn A, Lancaster P, Speirs A and Healy D (2000) Incidence of cancer in children born after in-vitro fertilization. Hum Reprod 15,604–607.[Abstract/Free Full Text]

Bryan E, Little J and Burn J (1987) Congenital anomalies in twins. Baillieres Clin Obstet Gynaecol 1,697–721.[ISI][Medline]

Daniel Y, Ochshorn Y, Fait G, Geva E, Bar-Am A and Lessing JB (2000) Analysis of 104 twin pregnancies conceived with assisted reproductive technologies and 193 spontaneously conceived twin pregnancies. Fertil Steril 74,683–869.[CrossRef][ISI][Medline]

Dhont M, Sutter PD, Ruyssinck G, Martens G and Bekaert A (1999) Perinatal outcome of pregnancies after assisted reproduction: a case-control study. Am J Obstet Gynecol 181,688–695.[ISI][Medline]

Ericson A and Källén B (2001) Congenital malformations in infants born after IVF: a population-based study. Hum Reprod 16,504–509.[Abstract/Free Full Text]

Fitzsimmons BP, Bebbington MW and Fluker MR (1998) Perinatal and neonatal outcomes in multiple gestations: assisted reproduction versus spontaneous conception. Am J Obstet Gynecol 179,1162–1167.[ISI][Medline]

Imaizumi Y (1998) A comparative study of twinning and triplet rates in 17 countries, 1972–1996. Acta Genet Med Gemellol 47,101–114.[Medline]

Källen B, Olausson PO and Nygren KG (2002) Neonatal outcome in pregnancies from ovarian stimulation. Obstet Gynecol 100,414–419.[Abstract/Free Full Text]

Knudsen L and Olsen J (1998) The Danish Medical Birth Registry. Dan Med Bull 45,320–323.[ISI][Medline]

Koivurova S, Hartikainen AL, Gissler M, Hemminki E, Sovio U and Jarvelin MR (2002) Neonatal outcome and congenital malformations in children born after in-vitro fertilization. Hum Reprod 17,1391–1398.[Abstract/Free Full Text]

Koudstaal J, Bruinse HW, Helmerhorst FM, Vermeiden JPW, Willemsen WNP and Visser GHA (2000) Obstetric outcome of twin pregnancies after in-vitro fertilization: a matched control study in four Dutch University hospitals. Hum Reprod 15,935–940.[Abstract/Free Full Text]

Lambalk CB and van Hooff M (2001) Natural versus induced twinning and pregnancy outcome: a Dutch nationwide survey of primiparous dizygotic twin deliveries. Fertil Steril 75,731–736.[CrossRef][ISI][Medline]

Larsen H, Nielsen GL, Bendsen J, Flint C, Olsen J and Sørensen HT (2003) Predictive value and completeness of the registration of congenital abnormalities in three Danish population-based registries. Scand J Public Health 31,12–16.[Medline]

Loos R, Derom C, Vlietinck R and Derom R (1998) The East Flanders Prospective Twin Survey: a population-based register. Twin Res 1,167–175.[Medline]

Nyboe Andersen A, Westergaard HB and Olsen J (1999) The Danish in vitro fertilisation (IVF) Registry. Dan Med Bull 46,357–360.[ISI][Medline]

Nygren KG and Nyboe Andersen A (2002) Assisted reproductive technology in Europe, 1999. Results generated from European registers by ESHRE. Hum Reprod 17,3260–3274.[Abstract/Free Full Text]

Olivennes F, Kadhel P, Rufat P, Fanchin R, Fernandez H and Frydman R (1996) Perinatal outcome of twin pregnancies obtained after in vitro fertilization: comparison with twin pregnancies obtained spontaneously or after ovarian stimulation. Fertil Steril 66,105–109.[ISI][Medline]

Pinborg A, Loft A, Schmidt L and Nyboe Andersen A (2003a) Attitudes of IVF/ICSI twin mothers towards twins and single embryo transfer. Hum Reprod 18,621–627.[Abstract/Free Full Text]

Pinborg A, Loft A, Schmidt L and Nyboe Andersen A (2003b) Morbidity in a Danish National Cohort of 472 IVF/ICSI twins, 1132 non-IVF/ICSI twins and 634 IVF/ICSI singletons: health-related and social implications for the children and their families. Hum Reprod 18,1234–1243.[Abstract/Free Full Text]

Pinborg A, Loft A, Schmidt L, Langhoff-Roos J and Nyboe Andersen A (2003c) Maternal risks and perinatal outcome in a Danish National cohort of 1005 twin pregnancies: The role of in vitro fertilisation (IVF). Acta Gynecol Obstet Scand (in press).

Rydhstroem H (1996) Pregnancy with stillbirth of both twins. Br J Obstet Gynaecol 103,25–32.[ISI][Medline]

Schachter M, Raziel A, Shevach F, Strassburger D, Bern O and Ron-El R (2001) Monozygotic twinning after assisted reproductive techniques: a phenomenon independent of micromanipulation. Hum Reprod 16,1264–1269.[Abstract/Free Full Text]

Schieve LA, Meikle SF, Ferre C, Peterson HB, Jeng G and Wilcox LS (2002) Low and very low birth weight in infants conceived with use of assisted reproductive technology. N Engl J Med 346,731–737.[Abstract/Free Full Text]

Sperling L and Tabor A (2001) Twin pregnancy: the role of ultrasound in management. Acta Obstet Gynecol Scand 80,287–299.[CrossRef][ISI][Medline]

Weinberg W (1902) Beiträge zur Physiologie und Pathologie der Mehrlingsgebuhrten beim Menschen. Arch Gesamte Physiol Menschen Tiere 88,346–430.

Westergaard HB, Johansen AMT, Erb K and Nyboe Andersen A (1999) Danish National In-Vitro Fertilization Registry 1994 and 1995: a controlled study of birth, malformations and cytogenetic findings. Hum Reprod 14,1896–1902.[Abstract/Free Full Text]

Submitted on June 6, 2003; resubmitted on October 2, 2003; accepted on October 9, 2003.