Prenatal Ultrasound and Autism: Is There a Connection?

BACKGROUND AND PURPOSE:

• Approximately 1/68 children are thought to be affected with autism spectrum disorder (ASD)
• Claims have been made that increasing use of prenatal ultrasound may be related to increasing rates of autism
• Rosman et al. (JAMA Pediatrics, 2018) studied whether there is an association between prenatal ultrasound exposure and the development of ASD

METHODS:

• Case-control study
• Population: Patients with ASD and control groups with developmental delay (without ASD) or typical development were identified retrospectively
• Ultrasonographic exposure was quantified by
  • Number and timing of scans
  • Duration of exposure
  • Mean strength (depth, frame rate, mechanical index, and thermal index)
  • Time of Doppler
  • 3- and 4-dimensional imaging
• Main outcomes
  • Ultrasound exposure and associated ASD risk were compared among the above groups, adjusting for infant sex, gestational age and maternal age
  • Ultrasound exposure was quantified and compared per trimester and entire pregnancy

RESULTS:

• Data were collected from 420 participants
  • 328 boys (78.1%); 92 girls (21.9%)
  • Mean age of follow up: 6.6 years
• ASD group
  • Received their first scan a mean of 10 days later than those with typical development
  • Were more likely to have received late prenatal care, and born to mothers ≥35 years of age
    • Socioeconomic factors may play a role, associated with increased risk for ASD and less access to prenatal care
  • Comparing ASD to developmental (non-ASD) delay group, there was no statistical difference in the timing of the initial scan, the number of scans or total exposure time
  • There was no statistical difference in the timing or duration of Doppler, 3-D, or 4-D studies in ASD group compared to the other 2 groups
  • The number of scans was not significantly different between groups
    • 5.9 scans in the ASD group (95% CI, 5.2-6.6),
    • 6.1 scans in the developmental delay group (95% CI, 5.4-6.8)
    • 6.3 scans in the typical development group (95% CI, 5.8-6.8)
  • Compared with the typical development group, the ASD group had shorter duration of ultrasound exposure during first and second trimester
    • First trimester: 290.4 seconds (95% CI 212.8-368.0 seconds) in the ASD group vs 406.4 seconds (95% CI, 349.5-463.3 seconds) in the typical development group
    • Second trimester: 1687.6 seconds (95% CI, 1493.8-1881.4 seconds) in the ASD group vs 2011.0 seconds (95% CI, 1868.9-2153.1 seconds) in the typical development group
  • ASD group had a greater mean depth of penetration compared to developmental delay group
    • First trimester: 12.5 cm (95% CI, 12.0-13.0 cm) in the ASD group vs 11.6 cm [95% CI, 11.1-12.1 cm) in the developmental delay group
  • ASD group had a greater mean depth of penetration compared to typical development group
    • First trimester: 12.5 cm (95% CI, 12.0-13.0 cm) in the ASD group vs 11.6 cm (95% CI, 11.3-12.0 cm) in the typical development group
    • Second trimester: 12.9 cm (95% CI, 12.6-13.3 cm) in the ASD group vs 12.5 cm (95% CI, 12.2-12.7 cm) in the typical development group

CONCLUSION:

• Mean number of scans was >5, higher than that typically recommended for normal pregnancies
• ASD group did not have increased number of scans, and had shorter duration of ultrasound exposure when compared to typical development group
• There was a significantly greater mean depth of ultrasonographic penetration in the ASD group compared with the developmental delay group in the first trimester and compared with the typical developmental group in the first and second trimester
• Study limitations noted by authors include
  • Retrospective
  • Study did not include preterm infants who are at higher risk for ASD
  • Accurate data on smoking, which is associated with ASD, was not available
  • Association between BMI, gestational diabetes and ultrasound depth requires further evaluation
• The companion editorial included the following discussion points
  • The increased depth of ultrasound may actually indicate less exposure “because ultrasonography attenuates as it propagates, and a larger propagation distance for children with ASD would mean less exposure in utero (all else being equal)”
  • Overall findings in this study do not support the concept that ultrasound is a single causal factor in ASD

Learn More – Primary Sources:

Association of Prenatal Ultrasonography and Autism Spectrum Disorder

Editorial: Prenatal Ultrasonography and the Incidence of Autism Spectrum Disorder

Prenatal Sonographic Features and Parental Origin of Triploidy

BACKGROUND AND PURPOSE:

• Triploidy results from an extra haploid chromosome set of paternal (diandric) or maternal (digynic) origin
• Occurs in 1–2% of all conceptions, with majority resulting in early embryonic/fetal loss
• McFadden and Kalousek prenatal sonographic assessment
  • Diandric phenotype: Relatively normal-sized fetus with microcephaly or normal head circumference, but abnormally large, cystic placenta
  • Digynic phenotype: Asymmetric fetal growth and non-cystic placenta
• Massalska et al. (Prenatal Diagnosis, 2017) sought to determine
  • The pattern of sonographic abnormalities in triploid pregnancies
  • The efficacy of McFadden and Kalousek prenatal sonographic assessment

METHODS:

Retrospective case analysis in a single unit between 1997 and 2015

RESULTS:

• Data analysis of 67 triploidy cases (64 singletons and 3 twins) between 11 and 30 weeks gestation
• All invasive testing was the result of fetal growth restriction and/or fetal anomalies
  • NT was increased in 25% of fetuses with diandric or undefined phenotype and no fetus with digynic phenotype, consistent with previous research
• Abnormal ultrasound findings included
  • Fetal growth restriction: 81.5%
  • Oligohydramnios: 50.7%
  • Structural defects: 61.2%
  • Multiple structural anomalies: 35.8%
    • Central nervous system, heart and urinary tract were most common
• ‘Diandric Phenotype’ was seen in 11.9% of fetuses that presented with cystic placentas
• ‘Digynic phenotype’ was identified in 70.2% asymmetrically growth-restricted fetuses with non-cystic placentas
• 17.9% of cases featured both diandric and digynic phenotype
• There were no live births in this cohort in cases with follow-up due to termination, IUFD or spontaneous miscarriage

CONCLUSION:

• Almost 40% of triploid fetuses do not appear to have structural abnormalities
• Diandric triploidy based on previous literature (small ‘n’) carries high risk for maternal complications
  • Preeclampsia and HELLP reported to be between 4 and 35%
  • Molar placenta changes not required
• In approximately 20% of triploid pregnancies, parental origin cannot be detected using McFadden and Kalousek assessment

Learn More – Primary Sources:

Triploidy – variability of sonographic phenotypes