Neurobehavioral Responses in Fetuses At Risk for Autism

Background: Research has begun to identify neurobehavioral differences in infants at risk for autism. In addition, genetic and biological findings point to prenatal influences on the development of autism. Research has not yet examined prenatal neurobehavioral alterations in infants at risk for autism. However, fetal behavioral and physiological responses in utero can be measured and may reflect the neurobehavioral status of fetuses at risk for autism.

Objectives: To investigate alterations in physiological and behavioral responses and characteristics in fetuses at risk for autism.

Methods: Pregnant mothers were recruited into two groups, those with a first degree relative with autism (Autism Risk group; AR) and those with no family history of autism (Low Risk; LR). Parent reported diagnoses in relatives were confirmed by best estimate clinical diagnosis by a clinical psychologist plus above-threshold scores on the ADOS. Twenty (20) participants have been recruited into this study, including 8 LR and 12 AR. Of the 12 who screened into the AR group, 4 had siblings with confirmed diagnoses. Fetal observations were conducted at 34 – 36 weeks gestational age during a 60 min fetal ultrasound (u/s) session. The session included a baseline period (10-15 min.), playback of a recording of the mother reading a story (2 min), an unfamiliar female reading the story (2 min), and single administration of a vibroacoustic stimulus. Each stimulus was followed by a 12-minute observation period. Fetal neurobehavioral measures were obtained by ultrasound and actocardiograph. Ultrasounds were conducted using a Toshiba diagnostic u/s machine (SSA-340A with a 3.75 MHz transducer) and a Toitu MT325 actocardiograph.  Data from the two machines was synchronized to facilitate later coding and analysis of fetal behavior and HR. Fetal behaviors were coded from digitized recordings using established protocols (Salisbury, 2005). Fetal movements (from actigraphy) and fetal HR were reviewed for artifacts and then summarized within each observational period. Analyses compared the 8 LR fetuses to the 4 AR fetuses with confirmed sibling diagnoses.

Results: The AR and LR fetuses did not differ in fetal heart rate (FHR) during an initial baseline period. But, AR fetuses did have significantly higher FHR than LR subjects in response to the mother’s voice (p = .012) and to stranger’s voice (p = .021). The groups did not differ in FHR response to a vibroacoustic stimulus. Groups did not differ on measures of FHR variability, but there was a trend for greater frequency of HR decelerations in the AR group. Behavioral coding indicated that the AR fetuses had higher stress signs as coded from u/s than did the LR subjects (3.4% vs. 0.4% across observation periods). This did not reach statistical significance given our small sample size, and appeared to be driven by one AR participant.

Conclusions: This small and preliminary study suggests that infants at risk for autism may show differences in neurobehavioral responses to in utero stimulation and experience. We will discuss the use of multilevel fetal observation for the study of the early developmental course in autism.