Abstract
Study question
Do in vitro fertilization (IVF) offspring present different neurodevelopment assessed by fetal neurosonography and infant neurobehavioral tests as compared to those spontaneously conceived (SC)?
Summary answer
IVF offspring, especially those obtained after fresh embryo-transfer (ET), showed subtle structural differences in fetal neurosonography and poorer neurobehavioral scores at twelve months of age.
What is known already
The number of pregnancies following assisted reproductive technologies (ART) is currently increasing worldwide. Concerns about the neurodevelopment of subjects conceived by IVF have been rising and mostly studied in children and adolescents with inconsistent results. Many of the identified risk associations were only observed in subgroups or disappeared after adjustment for covariates, mainly multiple pregnancy and gestational age at birth. It is unknown whether fetal brain development and cortical folding differ prenatally in IVF fetuses as compared to SC.
Study design, size, duration
This is the first study examining fetal neurodevelopment by neurosonography in IVF fetuses. A prospective cohort study of 210 singleton pregnancies recruited from 2017 to 2020, including 70 SC gestations, 70 conceived by IVF following frozen ET (FET) and 70 IVF after fresh ET. Fetal neurosonography was performed in all pregnancies. Additionally, Ages & Stages Questionnaires (ASQ) were obtained at 12 months of corrected age.
Participants/materials, setting, methods
IVF pregnancies were recruited from a single Assisted Reproduction Center, ensuring homogeneity in IVF stimulation protocols, endometrial preparation, laboratory procedures and embryo culture conditions. SC pregnancies were randomly selected from low-risk fertile couples and paired to IVF by maternal age. Fetal neurosonography including transvaginal approach was performed at 32±2 weeks of gestation, measured off-line by a single investigator and normalized by biparietal or occipitofrontal diameter. ASQ were obtained postnatally, at 12 months of corrected age.
Main results and the role of chance
Study groups were similar and comparable regarding maternal age, body mass index, study level and employment rate together with exposure to smoke, alcohol, aspirin and corticoids during pregnancy, gestational age (32±2 weeks) and estimated fetal weight (1700±400g) at neurosonography.
As compared to SC pregnancies, both IVF populations showed differences in cortical development with reduced parieto-occipital (fresh ET 12.5mm [SD 2.5] vs FET 13.4 [2.6] vs SC 13.4 [2.6]), cingulate (fresh ET 5.8 [IQR 4.2–7.4] vs FET 5.8 [4.1–7.5] vs SC 6.5 [4.8–7.8]) and calcarine (fresh ET 13.5 [IQR 10.1–16.1] vs FET 14.5 [12.1–15.8] vs SC 16.4 [14.3–17.9]) sulci depth together with lower Sylvian fissure grading. Cortical development changes were more pronounced in the fresh ET group as compared to FET. Corpus callosum length and insula depth were lower in FET and fresh ET groups, respectively. Neurosonographic changes remained statistically significant after adjustment by ethnicity, gender, gestational age and weight centile at scan.
IVF infants showed worse ASQ scores, especially in fresh ET for communication, personal-social, fine-motor and problem-solving skills. Gross-motor scores were significantly lower in FET as compared to SC and fresh ET. Differences were statistically significant after adjustment by maternal ethnicity, study level, employment status, breastfeeding, gender and corrected age.
Limitations, reasons for caution
The reported neurodevelopmental differences are subtle, with most neurosonographic findings lying within normal ranges. Infertility factors contribution to the outcome cannot be unraveled from the ART procedure itself. The milder features found in FET individuals cannot condition the techniqués choice and must be considered together with their global perinatal results.
Wider implications of the findings: Neurosonography is an appropriate tool to identify subtle brain differences between fetuses exposed and not exposed to ART. Prenatal features were consistent with postnatal neurobehavioral findings. These results support the relevance of a neurodevelopmental follow-up in IVF patients. Further studies are warranted to assess the long-term performance in these subjects.
Trial registration number
Not applicable
Calibrated Bayesian Neural Networks to Estimate Gestational Age and Its Uncertainty on Fetal Brain Ultrasound Images
