Abnormal Neonatal Auditory Brainstem Response and 4 Month Arousal-Modulated Attention Are Jointly Associated with Autism Severity Scores in Childhood in NICU Graduates

Background:  

Early behavioral abnormalities more prevalent in Neonatal Intensive Care Unit (NICU) graduates later diagnosed with Autism Spectrum Disorder (ASD) have been reported by our group (Karmel et al., 2010).  One of these was a visual preference for high rates of stimulation (8>3>1Hz.) when less aroused at four months post term age (PTA) (seen in our Arousal-Modulated-Attention (AMA) task); a preference found in newborns. These results suggested early problems with the attention/arousal system in ASD children; a system proposed to be modulated by brainstem function.  Accordingly, we examined the extent to which another measure of early brainstem development, the auditory brainstem evoked response (ABR), contributed to the AMA preference in predicting the later emergence of ASD behaviors in NICU graduates.

Objectives:  

To evaluate the contribution of both neonatal ABRs and 4 month PTA AMA preferences to later ASD behaviors in NICU graduates.

Methods:  

As neonates, three CNS injury groups were defined based on ABR and cranial ultrasound  assessments as described in Karmel et al. (2010): 1) No Detectable CNS Insult (n=27); 2) Abnormal ABRs only (n=28); and 3) Mild structural injury (n=22). Visual preference to pairs of checkerboard patterns flashing at 1, 3, or 8Hz were obtained when infants were more and less aroused at 4 months PTA. Type, and severity of autistic behaviors were obtained at a mean (SD) age of 3.4 (1.2) years based on parent PDD Behavior Inventories (PDDBI; Cohen and Sudhalter, 2005) and IQs were obtained around the same age using the Griffiths Mental Development Scales.  

Results:  

Visual preferences were highly correlated with 5 of 10 PDDBI domain scores (all p values <.001) and 3 of 3 composite scores (all p values <.002) but these associations were specific to the Abnormal ABR group. These significant effects were for measures of social communication (rs -.64 to -.73), social pragmatic problems (r=0.63), and arousal modulation problems (r=.59). The Social Discrepancy composite score (which correlates best with diagnosis) showed strong effects (r = -0.74, p<.000); the greater the preference for higher rates at 4 months, the greater the social deficit at 3 years.  By contrast, there were no significant correlations with visual preference for the other CNS groups or with performance IQ for any group. Correlational differences across groups were confirmed with separate-slopes GLMs.  Ten of 14 children later diagnosed with ASD were in the Abnormal ABR group; 7 of these had strong preferences for the fastest rates. By contrast, the Mild CNS group had 3 ASD cases and the No Insult group had 1 ASD child.

Conclusions:

These findings indicate that the joint occurrences of abnormal neonatal ABRs and preference for more stimulation at 4 months PTA are markers for the development of autistic behaviors in this population; both indices of brainstem maturation problems. Since each is related to autistic behaviors, it not known whether the effects are independent or are sequential. Since abnormal ABRs typically are related to CNS structural injury, the current finding for Group 2 may represent a different development mechanism rooted in very early fetal brainstem development.