Assisted reproductive technology (ART) treatments have been used for more than 30 years1. The first child conceived with in vitro fertilization (IVF) in the world was born in 1978, and the first one in Spain was born in 1984. Since then, more than 5 million individuals have been conceived with ART, amounting to 1.3%–6.1% of the total live births in Europe each year.2 The success rate of ART has been increasing since the early days of the field thanks to the experience accrued and scientific advances, but concerns remain about the potential impact in child health of conception through these means.

There is an ample body of evidence on the psychomotor development and cognitive skills of children conceived with ART, and to date most authors have not found statistically significant differences relative to the general population.3–7 However, the evidence regarding the incidence of autism spectrum disorder (ASD) is contradictory.8–11 Although the aetiology of ASD can still not be determined in most cases, epidemiological studies suggest that, in addition to the genetic factors that have already been identified, other factors such as the health of mothers before pregnancy, exposure to toxic substances, parental infertility problems and fertility treatments, among others, play an important role.12

In this context, the primary objective of our study was to determine whether there is an increased risk of neurodevelopmental disorders in children up to age 3 years conceived with ART compared to naturally conceived (NC) children, independently of maternal age, the number of foetuses and preterm birth. In addition, we performed an analysis to identify factors specific to ART potentially associated with disorders diagnosed in the group conceived with ART.

We designed an observational prospective matched cohort study to compare psychomotor development in children conceived with ART (a group we will from now on refer to as “ART children”) and NC children of similar characteristics. The study was approved by the Ethics Committee of the Hospital Clínico Universitario Virgen de la Arrixaca. The signing of an informed consent form was required for participation in the study, in adherence with the principles of the Declaration of Helsinki.

During the period under study, we evaluated 231 ART children and 208 NC children (N = 439; Fig. 2). The comparative analysis of the general characteristics of both cohorts did not detect any significant differences in mean maternal age, the proportion of multiple gestation, the mean gestational age at birth, family history or exposure to substances during gestation. We found a higher proportion of mothers that received folic acid supplementation before conception, of mothers who became pregnant after more than 1 year of trying and of female babies in the ART group, and a higher proportion of monochorionic twin pregnancy in the control group (Table 1).

Figure 2.

Flowchart of sample selection.

(0.47MB).

Having completed the 2 evaluations before age 3 years in ART children, we found that 9% had some type of neurodevelopmental disorder of unknown aetiology:

• ˗

  GDD in 3%.

• ˗

  Language disorder in 7.8%.

• ˗

  ASD in 3.4%.

All patients that received a diagnosis of GDD or ASD were further assessed with the fragile X DNA test and array comparative genomic hybridization (aCGH), and some underwent blood and urine tests for metabolic assessment. These tests only detected a disorder in 1 child conceived with ART: trisomy 47, XXY compatible with Klinefelter syndrome.

The comparative analysis did not identify any difference between the 2 groups in the frequency of diagnosis of neurodevelopmental disorders (Table 2).

We did find differences in the number of children that underwent assessment with aCGH (7.3 in the ART group vs 1.9% in the NC group; P = .026), mainly due to the association of neurodevelopmental disorders and/or congenital malformations with craniofacial dysmorphic features, which were more frequent in the ART group.

In cases in which additional tests were performed because the clinical manifestations matched those of a specific genetic disorder (assessment of methylation status at 14q32, 11p15, sequencing of the EFNB1 gene), the results were normal.

Other specific disorders were also diagnosed in 3 children in the ART group, and these patients were eliminated from the subsequent analysis because they could have neurodevelopmental disorders as a result of their underlying disease: 1 girl with Steinert syndrome inherited from the mother (antenatal diagnosis), 1 case of Beckwith-Wiedemann syndrome with IC2 hypomethylation and 1 girl with hypohidrotic ectodermal dysplasia carrier status.

The multivariate analysis of factors related to ART that could be associated with the neurodevelopmental disorders detected in the ART group revealed the following:

When it came to the presence of some form of neurodevelopmental disorder of unknown cause (9%), logistic regression did not detect an association with any ART-related factors, but it did detect an association with multiple gestation (OR, 3.19; 95% confidence interval [CI], 1.00–10.13; P = .049).

In the case of GDD (3%), we also did not find an association with any ART-related factor, but we found an association with a positive history of neonatal hypoglycaemia, defined as a blood glucose concentration of less than 25 mg/dL in the first 4 h of life and less than 35 mg/dL thereafter, as specified in the protocol of our hospital,19 and with infertility of unknown cause (Table 3).

We found evidence of language disorder in 7.8% of the ART group by age 3 years. Although the univariate analysis found an association between language disorder and a positive history of ovarian hyperstimulation, neonatal hypoglycaemia and FET, the logistic regression analysis only found an association with FET (Table 4) independent of preterm birth, maternal age and multiple gestation.

Autism spectrum disorder was diagnosed in 3.4% of ART children. The univariate analysis did not detect an association with any ART-related factor or with multiple pregnancy, preterm birth etc.

Our analysis did not reveal any statistically significant differences in psychomotor development through age 3 years between children conceived with ART and NC children, which was consistent with the majority of the evidence published to date. The increasing use of fertility treatments and their known association with a higher incidence of perinatal disease (mainly low birth weight and prematurity) raised concerns that these children could consequently suffer from neurologic problems more frequently. Some authors have also proposed the hypothesis that the use of ARTs could have a direct effect on neurodevelopment through epigenetic mechanisms at play in the early stages of embryonal development.20 These hypotheses notwithstanding, most studies conducted to date and the most recent systematic review on the subject21 have not found significant differences in psychomotor development (gross and fine motor skills, communication and language, problem solving and cognitive skills) between the ART and the NC populations.

We found a high proportion of ASD (3.4%) in the ART cohort in our study, but the multivariate analysis did not find differences between the ART and NC groups. A recent review and meta-analysis that included 11 studies and a total of 8 161 225 patients found a relative risk of ASD of 1.35 (95% CI, 1.09–1.68; P = .007) in ART children.22 However, the authors noted that further studies were necessary to take into account potential confounders such as maternal age, preterm birth, multiple gestation etc., as most studies did not analyse their potential impact. Our findings support this conclusion. Although the sample size may also have an impact on the results, our data suggest that the criteria used to select the control group (maternal age, type of pregnancy and gestational age at birth) should be considered the most important risk factors associated to the morbidity found in the population of ART children, as opposed to the use of ART itself. This supports the currently growing practice of single-embryo transfer,23,24 which minimises the frequency of multiple gestation and, as a secondary effect, the frequency of preterm birth, something that should be discussed with couples.

On the other hand, the analysis of ART-specific factors potentially associated with the neurodevelopmental disorders detected in the ART cohort only identified FET as a risk factor for language disorder that was independent of multiple gestation, preterm birth and maternal age, an association not previously described in the literature. Among the limitations of the study, we ought to highlight the young age of our sample and the fact that we did not ask participating couples whether either partner had had language delays in childhood. Given the known familial component in isolated cases of language disorder, this history could have an impact on the results. However, there is a scarcity of long-term longitudinal data on the FET population, and given the current generalised practice of vitrifying all embryos, performance of studies on the subject is of the essence, with particular emphasis on psychomotor development.25 We did not find an association of any neurodevelopmental disorder with the type of ART used (IVF or ICSI), nor with the different stimulation protocols or the duration of embryo culture, contrary to the findings of other authors.26

We ought to highlight the association we found between infertility of unknown cause and GDD, which supports previous descriptions in the literature of an association between subfertility in couples and poorer psychomotor development outcomes and coordination in their children.27,28 The results of these studies evince that difficulties in conception have a more relevant impact on neurodevelopment that the ARTs that these couples may come to use.

This is the first prospective cohort study analysing the psychomotor development of children conceived with ART through age 3 years in comparison with a control group of similar characteristics in terms of multiple pregnancy, maternal age and preterm birth. We did not find a difference between groups in the incidence of the neurodevelopmental disorders under study through age 3 years, which was consistent with most of the previous literature. Our findings suggest that the criteria used for selection of control groups and subfertility should be considered potential risk factors, as opposed to ART itself. These results should be conveyed to couples planning to attempt conception with ART to improve counselling and promote elective single-embryo transfer, thus reducing the risk of multiple pregnancy and preterm birth, which are 2 known risk factors for neurodevelopmental abnormalities. The association between FET and language disorder has not been described in the past. Long-term longitudinal studies on the health of children conceived by means of FET are needed to confirm this finding.

The authors have no conflicts of interest to declare.