Mission & Constitution
Basic Science and Instrumentation
General and Abdominal Ultrasound
High Frequency Clinical and Preclinical Imaging
Ultrasound in Global Health
Ultrasound in Medical Education
AIUM Courses & Events
Accreditation & Disclosures
Ultrasound in Med Ed Portal
Ultrasound Awareness (MUAM)
Member Directory (Members Only)
Frequently Asked Questions
Access Accreditation Account
Accredited Practice Resources
Directory of Accredited Practices
Endowment for Education & Research
AIUM Responds to Autism Study
September 7, 2016
AIUM Bioeffects Committee Statement on Paper
Webb SJ, Garrison MM, Bernier R, McClinic AM, King BH, Mourad PD.
Severity of ASD symptoms and their correlation with the presence of copy number variations and exposure to first trimester ultrasound.
Autism Res 2016. doi:10.1002/aur.1690.
proposed that multifactorial disorders, such as autism spectrum disorders (ASDs), have a higher incidence when the brain of a genetically at-risk individual is exposed to an exogenous stressor during a critical time of neurodevelopment. Due to a lack of appropriate controls, Webb et al studied a modified version of this hypothesis, which was that a relationship exists between the severity of ASD and ultrasound exposure during the first trimester of pregnancy in fetuses with a genetic predisposition to ASD. The study reports on a retrospective analysis of data collected from families with 1 child with ASD aged 4 to 18 years (Simon’s Simplex Collection). The severity of ASD was characterized by measures of cognitive ability, social ability, and repetitive behaviors. Genetic predisposition was characterized by the presence of ASD-associated copy number variations (CNVs). Information on ultrasound exposure and the medical history was collected by interview.
Children with ASD and exposure to ultrasound in the first trimester showed a statistically significant decrease in 1 metric of observed social affective behavior symptoms but a statistically significant increase in 1 metric of parent-reported repetitive behaviors compared to children with no ultrasound exposure. However, there were no statistically significant differences in 2 other metrics of cognitive ability (Vineland Adaptive Behavior Scale II [VABS-II] and verbal IQ), 3 metrics of social affective behaviors (Standardized Autism Diagnostic Observation Schedule, Social Affect [ADOS SA]; Autism Diagnostic Interview–Revised, Spectrum Disorders [ADI-R SD]; and Social Responsiveness Scale [SRS]), and 3 metrics of repetitive behaviors (ADOS Repetitive Domain score [ADOS REP]; ADI-R, restrictive and repetitive behaviors [RRB]; and Repetitive Behavior Scale–Revised [RBS-R]). Children with ASD, an ASD-associated CNV, and exposure to first-trimester ultrasound showed a statistically significant decrease in nonverbal IQ compared to children with no ultrasound. However, again, there were no statistically significant differences in 2 other metrics of cognitive ability, 3 metrics of social behaviors, and 3 metrics of repetitive behaviors. Male children with ASD, an ASD-associated CNV, and exposure to first-trimester ultrasound showed statistically significant decreases in nonverbal IQ and statistically significant increases in repetitive behaviors compared to children with no ultrasound exposure. However, there were no statistically significant differences in 2 other metrics of cognitive ability, 3 metrics of social behavior, and 2 other metrics of repetitive behaviors.
The study did not test the original hypothesis because it did not include a comparison control group, namely, a sibling comparison group, and this was a significant flaw in the study. The occurrence of ultrasound in the first trimester was determined solely by parental recall. No information was provided on the details of the ultrasound exposure, including the indication for the ultrasound, type of ultrasound (abdominal or transvaginal), exposure duration, exposure during other trimesters, or imaging modality. Furthermore, the study did not test for numerous possible confounding factors. The authors also did not perform a Bonferroni correction for multiple comparisons, which would be appropriate for a confirmatory study establishing causal connections. The results are not significant at the 5% level with a Bonferroni correction. In the absence of such a correction, the authors should have explicitly stated that this was an exploratory analysis designed to uncover correlations that need to be confirmed in a future study. In summary, the results of the study do not demonstrate a causal link between ultrasound and autism. Ultrasound examinations can be safely performed by educated and credentialed sonographers and clinicians when medically indicated and when the ALARA (as low as reasonable achievable) principle is followed. The review article by Abramowicz
provides additional information on ultrasound and autism.
1. Casanova MF.
The neuropathology of autism.
Brain Pathol 2007; 17:422–433.
2. Abramowicz JS.
Ultrasound and autism: association, link, or coincidence?
J Ultrasound Med 2012; 31:1261–1269.