Poster #1 - Identifying Risk for Autism Spectrum Disorder Using Newborn Hearing Screening

• In Event: 3-001 - Poster Session 2 during Breakfast

Abstract

Autism spectrum disorder (ASD) refers to a group of developmental disorders characterized by social and communication problems, repetitive behaviors, and limited interests. The prevalence of ASD is 1 in 68 children and the median age of diagnosis is approximately 4 years. Although behavioral signs of ASD are present between 6-12 months, diagnosis is not typically considered reliable before 2 years of age. Earlier identification and intervention are critical to improving child outcomes and decreasing economic costs associated with ASD.

Auditory brainstem response (ABR) is a promising and feasible biomarker for early identification of ASD risk. ABR is a quick, noninvasive test routinely used to screen for hearing impairment in newborns around the world. Previous studies identified prolongation of ABRs in infants later diagnosed with ASD (for a review see Miron, Beam, & Kohane, 2017). These preliminary studies, while promising, had small sample sizes and replication is needed. Secondary datasets provide a promising opportunity to examine the predictive accuracy of ABRs on larger samples of children.

The present study linked secondary datasets to examine the potential of ABR as a biomarker for ASD risk. We integrated newborn ABR hearing screening results provided by MEDNAX with ASD diagnosis provided by the Center for Autism and Related Disabilities for one Florida county. In a sample of 21 children with ASD, wave V latency was significantly prolonged in the ASD group compared to the control group (n=8,490) for the left ear (ASD: mean= 8.94 milliseconds (ms), SD= 0.9; Control= 8.41 ms, SD=0.7; Cohen’s d=0.69, t(20)=2.6, p=0.018) but not the right ear.

We are now replicating and expanding this study with a larger sample (18 Florida counties) combining newborn ABR hearing screening results provided by MEDNAX with developmental disability classification for preschool-aged children from the Children’s Registry and Information System provided by the Florida Department of Education. Data integration is currently underway. We anticipate having a linked sample that contains the ABR results for several hundred children with ASD. The ABR wave V latency for left and right ears will be compared between the ASD group and the remaining newborns who will serve as the control group. We will examine mean group differences and predictive accuracy.

Data from an existing newborn screening test has the potential to provide neurological information that could identify newborns at the greatest risk for ASD and, in conjunction with early behavioral signs, be used to diagnose ASD at younger ages.

Authors

• Christine Delgado, University of Miami
  Presenting Author

• Elizabeth Simpson, University of Miami

• Oren Miron, Harvard Medical School

• Anibal Gutierrez, University of Miami

• Jillian Gerstenberger, MEDNAX, Inc.

• Rafael Delgado, University of Miami

• Isaac Kohane, Harvard Medical School