Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
To systematically review and, when feasible, pool, published data regarding the prevalence of childhood neurodevelopmental delay in fetuses with increased first-trimester nuchal translucency (NT), normal karyotype and absence of structural defects or identifiable syndromes.
MEDLINE and SCOPUS searches using combinations of the terms ‘nuchal translucency’ AND ‘outcome*’ were complemented by perusal of the references of the retrieved articles and an additional automated search using the ‘search for related articles’ PubMed function. Only children with a normal karyotype and no structural defects or syndromic abnormalities were included in the analysis. Between-studies heterogeneity was assessed using the I2 statistic.
The total prevalence of developmental delay in all 17 studies was 28/2458 (1.14%; 95% CI, 0.79–1.64; I2 = 57.6%). Eight studies (n = 1567) used NT > 99th centile as the cut-off; 15 children (0.96%; 95% CI, 0.58–1.58%) were reported as having developmental delay (I2 = 72.2%). Four studies (n = 669) used the 95th centile as the cut-off for increased NT; seven children (1.05%; 95% CI, 0.51–4.88%) were reported as having developmental delay (I2 = 29.2%). Five studies used 3.0 mm as the cut-off for increased NT; the pooled rate of developmental delay was six of 222 children (2.70%; 95% CI, 1.24–5.77%; I2 = 0.0%).
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Increased nuchal translucency (NT) is a common finding as, by definition, 5% of fetuses will have increased measurements > 95th centile. A large body of evidence has shown that increasing measurements of NT are associated with an increased risk for chromosomal abnormalities, genetic syndromes, structural abnormalities (mainly congenital heart defects), intrauterine infection and fetal demise; with NT measurements of > 6.5 mm, the chance of a liveborn and healthy baby is as low as 15%1, 2.
The long-term neurodevelopmental outcome of children who had increased NT as fetuses has recently attracted attention. Many of the conditions associated with increased first-trimester NT are themselves accompanied by neurodevelopmental delay. However, much less is known about the neurodevelopmental outcome of children with increased NT and a normal karyotype that were apparently healthy at birth (i.e. did not have structural abnormalities of recognizable genetic syndromes). Most of these data are circumstantial, being scattered across reports focusing generally on the outcome of fetuses with increased NT and a normal karyotype, without using formal tools for neurodevelopmental assessment.
Our aim was to systematically review evidence on the neurodevelopmental outcome of fetuses with increased first-trimester NT and no chromosomal, structural or recognizable syndromic conditions at birth.
We searched the literature (last update September 2011) for studies with measurement of NT during the first trimester of pregnancy and postnatal follow-up of these children. MEDLINE and SCOPUS searches used combinations of the terms ‘nuchal translucency’ AND ‘outcome*’. We deliberately used wide search criteria, as the relevant evidence is sparse. These searches were complemented by perusal of the references of the retrieved articles and an additional automated search using the ‘search for related articles’ PubMed function. All studies were carefully compared to ensure avoidance of duplicate or overlapping samples. In case of overlap, the study with the largest number of cases was included.
For the structure of the review we followed the methodology described by Stroup et al.3 and Whittemore and Knafl4.
Inclusion criteria were: cohort or case–control studies with prenatal measurement of NT during the first trimester and postnatal follow up; provision of sufficient details in order to identify only cases with isolated increased NT (i.e. without identifiable chromosomal abnormalities, syndromes or congenital structural defects); there were no language restrictions.
Cases were excluded in which increased NT was associated with chromosomal, genetic or structural abnormalities. The quality of each study was assessed according to (1) the ascertainment of neurodevelopment (formal assessment with standardized and validated tools vs. unstructured assessment e.g. telephone contact, interviews or non-standardized questionnaires, or not stated) and (2) blinding of the examiner assessing development to NT status (yes vs. no or not stated). A further quality assessment was carried out using the Newcastle–Ottawa scale for non-randomized studies5, which is based on the selection of the study population or populations, the comparability of the different groups participating and the ascertainment of either the exposure or the outcome of interest for case–control or cohort studies, respectively.
Data extraction and study quality assessment were independently performed by two authors (A.S. and S.P.); in the event of disagreement a consensus was reached after discussion between the two authors or after evaluation by the third author (G.M.).
The outcome measure was the rate of neurodevelopmental delay in children with increased first-trimester NT, normal karyotype and absence of structural abnormalities and identifiable syndromes. The studies were grouped according to the cut-off used for increased NT (all vs. > 99th centile vs. > 95th centile vs. 3 mm). The rates of developmental delay were calculated separately (1) for children with increased NT, a normal karyotype, a normal second-trimester anomaly scan and absence of structural defects and identifiable syndromes after birth, and (2) for children with increased NT, a normal karyotype and a normal second-trimester anomaly scan (i.e. before confirmation of normal anatomy at birth).
A secondary analysis was also attempted according to the presence or absence of nuchal edema at the second-trimester scan.
Descriptive statistics were given as proportions, together with their 95% CIs. Between-studies heterogeneity was assessed using the I2 statistic, which describes the percentage of variation that is present as a result of heterogeneity rather than chance. A simplistic grouping would tentatively assign adjectives of low, moderate and high to I2 values of 25%, 50% and 75%6. MetaDiSc 1.47 was used for pooling the results. Random effects models (REM) were applied in pooling the data.
Of the 337 items retrieved from our electronic search, 181 were excluded based on the title and the abstract. The remaining 156 articles were retrieved for screening of full text. We further excluded 79 studies that examined NT measurements and outcomes related to chromosomal abnormalities and excluded 53 that examined outcomes other than neurodevelopmental delay. We also excluded seven studies that assessed, genetically, fetuses with increased NT for the identification of a genetic syndrome with structural chromosomal abnormalities. In total, 17 studies provided data on the postnatal development of fetuses with isolated increased NT (Figure 1).
Descriptive characteristics of the studies that were included in the review are presented in Table 1. Details about the operators and the ultrasound machines used are shown in Table S1. The quality assessment of the studies according to the Newcastle–Ottawa scale is presented in Table S2: 13 were prospective or retrospective cohorts of fetuses with increased NT1, 8–19, and four were case–control, comparing fetuses with NT > 99th centile vs. fetuses with NT < 99th centile20, fetuses with NT > 95th centile vs. < 95th centile21, and fetuses with NT > 99th centile vs. fetuses with normal NT (< 95th centile)22, 23. Senat et al. examined prospectively fetuses originating from an unselected population but because of the anticipation of a high loss-to-follow-up rate, they compared the study population with a control group comprising 370 term children from a French national population-based cohort study designed in 199722. However, they compared individual developmental milestones between the two groups and did not give specific numbers for the prevalence of delay in the control group. The majority of studies did not include a structured assessment of neurodevelopment. The Ages and Stages Questionnaire (ASQ) with further clinical assessment of neurodevelopment in the event of abnormal results was used in three studies8, 22, 23; the Bayley Infant Neurodevelopmental Screen or Scale or Infant Development (BINS/BSID) was used in two studies9, 10; assessment of development by clinical pediatric examination was reported in two studies11, 22, while in the rest of the studies the information was retrieved from contact with parents or information from medical records or GPs. None of the studies was blinded.
Table 1. Description of studies included in the review
Definition of developmental delay
Methods for developmental assessment
Lost to follow-up
Including one case of neonatal death. %ile, percentile; IUD, intrauterine death; NT, nuchal translucency thickness; TOP, termination of pregnancy.
A score obtained by forms returned by the parents, the hospitals or the pediatricians
Forms returned by the parents or from the hospital, midwife reports, and from the local pediatricians. Also, they interviewed all parents on the telephone to assess postnatal growth, psychomotor skills and speech as well as interaction with the child
Failure to achieve normal developmental milestones
Assessment by pediatricians using the Mental Development Index (MDI) < 68 or the Psychomotor Development Index (PDI) < 68 (parts of Bayley BINS). Clinical follow-up, or general practitioners performed long-term follow-up by telephone investigations
Hospital-based assessment by Pediatrician or Ages and Stages Questionnaire (ASQ) from parents. If the child's score was below − 2 SD, a formal clinical neurodevelopmental assessment was performed by a pediatrician
Cases: fetuses with increased NT from an unselected population; controls: control group from a previous study
The Ages and Stages Questionnaires (ASQ): the screening cut-off for each domain was 2 SD below the mean score. If the child's score was below 2 SD in one or more domains, further assessment of the child's performance was recommended
All children were examined by a pediatrician within 2 days after birth and then at 1, 4, and 9 months, and at 2 years of age. This was completed by serial questionnaire to be answered by the parents
Within 2 days after birth and then at 1, 4 and 9 months, and at 2 years of age
Questionnaire sent to the parents (appears in the manuscript) and/or reviews of available medical records
33.5 (range, 7–75) months
From the 17 studies, only one14 referred to alcohol or drug abuse or to other exposures associated with impaired neurodevelopment, so an additional analysis was not feasible.
The 99th centile (or 3.5 mm) was used as the cut-off for increased NT in eight studies1, 8, 10, 12, 13, 20, 22, 23, the cut-off was 3 mm in five studies9, 11, 14–16 and four studies used the 95th centile as the cut-off17–19, 21.
Fetuses with increased first-trimester NT, a normal karyotype and absence of structural abnormalities or identifiable syndromes at birth
This analysis included liveborn fetuses with increased first-trimester NT, a normal karyotype and absence of structural abnormalities or identifiable syndromes at birth.
The total prevalence of developmental delay in all 17 studies was 28/2458 (1.14%; 95% CI, 0.79–1.64%; I2 = 57.6%).
In the eight studies1, 8, 10, 12, 13, 20, 22, 23 in which NT > 99th centile was used as the cut-off (n = 1567), 15 children (0.96%; 95% CI, 0.58–1.58%), one of which had an unidentifiable genetic syndrome, were reported as having developmental delay (I2 = 72.2%) (Figure 2). The pooled rate of delay was four of 317 (1.25%; 95% CI, 0.49–3.20%) in the three studies that used structured tools for screening for delay (I2 = 53.9%)8, 22, 23. In the two studies in which comparative information for rates of delay were provided for fetuses with and without increased NT, the rate of childhood developmental delay did not differ between fetuses with NT > 99th centile vs. those with NT < 95th centile (none of 80 vs. none of 137)23, or between fetuses with NT > 99th centile vs. those with NT < 99th centile (one of 80 vs. one of 297; difference = 0.22%; 95% CI, − 1.39 to 2.76%)20.
In the four studies17–19, 21 in which the 95th centile was used as the cut-off for increased NT (n = 669), seven children (1.05%; 95% CI, 0.51–4.88%), three of which had an unspecified genetic syndrome, were reported as having developmental delay (I2 = 29.2%) (Figure 3).
In the five studies9, 11, 14–16 in which 3.0 mm was used as the cut-off for increased NT, the pooled rate of developmental delay was six of 222 (2.70%; 95% CI, 1.24–5.77%; I2 = 0.0%) (Figure 4). One of these children was reported as having an unspecified genetic syndrome.
Fetuses with increased first-trimester NT, a normal karyotype and a normal second-trimester scan
This analysis included liveborn fetuses with an increased first-trimester NT, a normal karyotype and a normal second-trimester anomaly scan. Its difference from the previous analysis was that it also included fetuses in which abnormalities were identified after birth.
In the eight studies1, 8, 10, 12, 13, 20, 22, 23 (n = 1666) in which the 99th centile was used as the cut-off for increased NT, the rate of developmental delay was 15/1666 (0.90%; 95% CI, 0.55–1.48%).
In the four studies17–19, 21 (n = 691) that used the 95th centile as the cut-off for increased NT, there was a 1.01% (seven of 691 fetuses; 95% CI, 0.49–2.07%) pooled rate of developmental delay.
In the five studies9, 11, 14–16 that used 3.0 mm as the cut-off for increased NT, there was a 2.47% (six of 243 fetuses; 95% CI, 1.14–5.28%) pooled rate for developmental delay.
Presence of nuchal edema at the second-trimester scan
Four studies1, 8, 17, 20 provided adequate data in order to construct 2 × 2 tables for the neurodevelopment of infants according to the persistence or absence of second-trimester nuchal edema. After excluding cases with chromosomal abnormalities, structural defects, prenatal or perinatal death and identifiable syndromes, the pooled rate of childhood neurodevelopmental delay in fetuses with NT > 99th centile and no second-trimester nuchal edema was 10/1494 (0.66%, 95% CI 0.36–1.21), vs. 1/94 (1.06%, 95% CI 0.19–5.78) for fetuses with persistent nuchal edema1, 8, 20.
We have found that the pooled rate for neurodevelopmental delay in studies of fetuses with increased first-trimester NT, normal fetal karyotype and lack of structural defects or identifiable syndromes is in the region of 1%, both for the overall population of fetuses characterized as having increased NT and for the more meaningful cut-off of NT > 99th centile. This rate is, at worse, no higher than the rates reported for the general population, which lie in the region of approximately 3%24–26.
Increased first-trimester NT has long been recognized as a risk factor for chromosomal abnormalities, fetal structural defects (mostly congenital heart defects), genetic syndromes and prenatal or perinatal demise27. Many of the syndromes associated with increased NT have developmental delay as a clinical feature, and, in turn, an underlying etiology (e.g. chromosomal, genetic, metabolic or structural abnormalities) can be found in approximately half (55–60%) of the children with global or motor delay28, 29. So, although the association between increased NT and fetal abnormalities (structural, chromosomal or syndromic) is well established, much less is known about the neurodevelopment of apparently healthy children with increased NT, and this is the group we focused on. We have to stress here the term ‘apparently’, as published studies did not provide data on how laborious and time consuming was the diagnostic workup in cases of developmental delay or disorder, and therefore some of their cases may actually be syndromic.
The low rate for developmental delay did not differ across the subgroup analyses we performed. Thus, there was no significant difference between fetuses with a normal second-trimester scan and those which were found to be normal at birth. This is explained by the facts that (1) the detailed anomaly scan has a high sensitivity for detecting structural defects and (2) the majority of non-syndromic cases have normal anatomy anyway. Therefore, pregnant women with a fetus with increased NT and a normal karyotype can be reasonably reassured of its developmental outcome after a normal second-trimester scan.
We have also found that the pooled rate of developmental delay did not differ according to the cut-off used to define increased NT (95th centile, 99th centile and 3.0 mm). In order to explore this further, we identified the individual NT measurements (when available) of fetuses with neurodevelopmental delay. The most suitable population for this was the pool of fetuses with NT > 95th centile, but unfortunately there were only five cases with known NT measurements. The NT was > 3.5 mm in four (80%); this may indicate an over-representation given that NT measurements of > 3.5 mm comprise only one fifth of NTs > 95th centile, but the sample was too small to draw any conclusions. For the more skewed population of fetuses with NT > 99th centile, the median NT for cases with neurodevelopmental delay was 4.0 mm (individual measurements were available for 10 fetuses). Finally, our secondary analysis did not indicate significant differences between fetuses with persistent or absent nuchal edema in the second trimester; it should be highlighted, however, that the group of fetuses with nuchal edema was small and the 95% CIs were wide. Intuitively, we were expecting to find a stronger association between higher NT measurements and developmental delay if NT was associated with delay at all. If, on the other hand, increased NT is not associated with delay, then it is reasonable that cases with delay are randomly scattered across different NT levels.
The results of our review should be interpreted with caution in general, as the data have further limitations.
First, there is no firm consensus on the definition of developmental delay. According to the American Academy of Pediatrics, developmental delay refers to the condition in which a child is not developing and/or achieving skills in the expected time frame, as opposed to developmental disorder (or disability), which refer to a childhood mental or physical impairment, or combinations of those, resulting in substantial functional limitations in major life activities30. However, the two terms are sometimes used interchangeably, and the term ‘delay’ is often used to describe mental retardation31. The studies included in our review appear to follow the definition of delay in achieving developmental milestones; therefore, many of the cases with delay are not expected to develop a permanent handicap.
Second, screening for, and ascertainment of, developmental delay can be problematic. Most of the studies used telephone interviews with parents in order to screen for delay. Asking the parents simple questions about their child's behavior and development may elicit quality information, but parents may be underestimating or overestimating their concerns about the development of their children26, 30. The most structured tools for developmental screening in the studies of our review were the ASQ, which has sensitivity and specificity in the region of 70–90%32 and was used in three studies8, 22, 23, and the BINS30, 33, with a similar performance to the ASQ, which was used in two studies9, 10. Both of these tools are screening tests, which means than in cases of high-risk score, a more detailed neurodevelopmental assessment is warranted. Only the two most recent studies8, 23 give detailed data on both abnormal ASQ and clinically assessed delay; 30 infants had abnormal ASQ scores and only four (13%) were eventually diagnosed with delay (one had del22q11 diagnosed at 2 years of age). In the rest of the studies it is not clear whether the diagnosis of delay refers to high-risk results at screening or to formal diagnosis after clinical assessment. Therefore, it appears that there are two possible sources of bias: screening bias, resulting from the suboptimal methods to identify children at risk, and ascertainment bias, resulting from insufficient information about the steps which followed identification of high-risk infants at screening. This problem is partially solved in case–control studies, when the same methodology is applied to both groups; the four case–control studies in our review20–23 did not report different rates of developmental delay between the two groups.
The different definitions, sample sizes, screening and ascertainment methods for developmental delay may also be responsible for the heterogeneity between studies, which was profound in some instances.
Third, it is not certain that syndromes can always be reliably identified. At least five of the 28 cases reported in our review had an unidentified genetic syndrome, of which developmental delay is a part, and there are probably even more cases not recognized as such. Three of the more recent studies reported supplementary chromosomal analysis using high-resolution comparative genomic hybridization (HR-CGH)10 or multiplex ligation-dependent probe amplification (MLPA)8, 10, 23 either for their whole sample or when a syndrome was suspected, and it appeared that these methods did not increase the detection of syndromes. We chose to include these cases for pragmatic reasons; if these conditions cannot be identified as specific syndromes prenatally or postnatally, from the viewpoint of prenatal counseling they are grouped together with the other apparently isolated cases.
Still, despite all these potential sources of bias, it is highly unlikely that fetuses with apparently isolated increased NT have an increased risk for developmental delay. In fact, many of the bias sources (e.g. the preselected nature of the ‘high-risk’ infant population, or the inclusion of syndromic cases which would increase the numerator) would err in the direction of an increased rate of delay in children with increased NT.
In conclusion, it appears that the risk for developmental delay in fetuses with increased first-trimester NT is not increased compared with the general population, after chromosomal abnormalities, structural defects and genetic syndromes are excluded. With the exception of a minority of studies focusing on this topic, most of the available data are scattered across studies reporting in general on the outcome of children with increased NT and fetal karyotype, which, at last partly, explains the inconsistency of methodology and results among different studies.
SUPPORTING INFORMATION ON THE INTERNET
The following supporting information may be found in the online version of this article:
Table S1 Technical characteristics of the ultrasound machines used and operators performing the assessments.
Table S2 The Newcastle–Ottawa scale for quality assessment of the included studies.