Very Early Brainstem Function Together with Attention Regulation Relate to Later ASD in NICU Graduates: Replication and Extension

Background:  

Identifying early indicators for Autism Spectrum Disorders (ASDs) could be useful for screening. We have reported behavioral abnormalities more prevalent in NICU graduates later diagnosed with ASD (Karmel et al, 2010), including visual preferences for higher stimulation rates when less aroused at 4 months post-term-age (PTA), an attention pattern more typical of newborns. Such dysfunction suggests problems in early arousal-modulated-attention (AMA), likely regulated by brainstem processes. We further have reported conjoint association of this abnormal attention regulation with abnormal transmission speeds in auditory brainstem evoked responses (ABRs), also reflecting brainstem dysfunction, predicts increased ASD risk in NICU graduates (Cohen et al, AS 2012). We now report on different samples and relations to later behavioral development.          

Objectives:  

To ascertain if previous findings: (1) can be replicated in an independent subsequently-diagnosed sample; (2) can be generalized to non-ASD samples at risk for ASD such as younger siblings of ASD children and high-medical-risk NICU infants, and (3) are related to atypical development associated with later-emerging ASD such as neurobehavioral functioning on standardized tests.   

Methods:  

Medically at-risk infants and infant siblings recruited as newborns for longitudinal follow-up studies were compared. Four groups defined: 1. ASD(Dx1): initial sample, n =27; 2. ASD(Dx2): replication sample, n=21; 3. non-ASD(SIB): younger siblings of ASD, n =36; 4. non-ASD(OTHER): remaining NICU cohort, n=1740. Groups contrasted on: (1) ABR neural transmission speeds: click ABRs obtained within a few days after birth (after 32wks post-conception). Abnormality was determined by component peak latencies significantly deviating from age norms at 80dBnHL;~12Hz for Waves I,III,V and III-V interval, which typically resolved before hospital discharge; (2) Attention regulation: AMA measured at 4 months PTA in paired-comparison design with 2 same checkerboard-patterned stimuli flashing at 1,3, or 8 Hz square-wave modulated frequencies when infants more and less aroused due to feeding and additional stimulation. Slope of the preference function across the 3 stimuli > .1 defined looking at higher rates (8Hz>3Hz>1Hz); (3) Standardized measures: Bayley Infant Neurobehavioral Screener (BINS), 2½mos; Bayley Scales of Infant Development (BSID), 4-7mos.     

Results:  

There was a very high incidence of abnormal ABRs related to ASD compared to the overall NICU sample (72% vs. ~28%), with non-ASD(SIB) ~50%. The 2 diagnosed samples did not differ in % abnormal ABRs or any behavioral measures indicating replication and were combined (ASD(DxC)) for analyses. ASD(DxC) showed more looking at higher rates than non-ASD(OTHER). The combined effect of an initial abnormal ABR with AMA slope >.1 predicted ASD. 60% of ASD(DxC) had both compared to 17% non-ASD(OTHER) and 0% non-ASD(SIB). Interestingly, non-ASD(SIB) had significantly lower scores on the BINS compared to ASD(DxC) and non-ASD(OTHER), which  may represent developmental problems not within criteria for ASD, although no differences on BSID were apparent at these young ages and small sample sizes.            

Conclusions:  

Dysfunction during early brainstem development indicated by altered neuronal transmission speeds, along with deficits in attention regulation as early as 4 months of age, may represent a potentially important path leading to later ASD not apparent through standardized tests during early infancy.