Diminished Social Responses During Neurobehavioral Exam in Newborns At-Risk for Autism

Background: Diminished social responses and initiations have been observed late in the first year of life in infants at risk for autism. Findings of altered social responses earlier in infancy have been less consistent. Observational methods that include specific presses for responses to social (animate) and nonsocial (inanimate) stimuli may be sensitive to such differences in social responsiveness in early infancy.   

Objectives:  To investigate differences in responses to animate and inanimate stimuli in the newborn period for infants at risk for autism. A secondary goal was to explore other behavioral differences on neurobehavioral exam in at risk infants.

Methods:  Nine (9) newborns were recruited from a study of infant siblings with autism (Autism Risk group; AR). Sibling diagnoses were confirmed for 6 of the children by best estimate clinical diagnosis plus above-threshold scores on the ADOS. These 6 AR infants were compared to a group of 45 newborns with no family history of autism recruited from an ongoing longitudinal study (Low Risk group; LR). All babies were administered the NICU Network Neurobehavioral Scales (NNNS; Lester & Tronick, 2004) at 2 to 4 weeks of age. The NNNS includes 6 procedures that press the baby to attend to, orient toward or visually track animate (examiner’s voice and/or face) and inanimate (rattle/ball) stimuli (Animate auditory, visual and auditory+visual stimuli; Inanimate auditory, visual and auditory+visual stimuli). The NNNS examiner was naïve to risk group status. Statistical analyses utilized one-tailed T-tests of means on the a priori hypothesis that the AR group would show diminished responses to the animate stimuli. Summary scales overall are also available from the NNNS and have been explored on a post hoc basis.

Results: The Animate visual and auditory items were averaged, as were the Inanimate items. This resulted in summary variables for Inanimate and Animate responses. The AR newborns showed diminished responses as compared to LR infants to the Animate items (AR 3.3 vs LR 5.2; p= 0.003; d= -1.3). The AR and LR infants did not differ on the Inanimate summary measure (AR 5.0 vs LR 5.2, p= 0.79; d= -0.2). The difference in Animate responses was driven by statistically significant differences on Animate Visual (p = 0.002) and the Animate Auditory+Visual items (p= 0.005). There was a trend for diminished responses to Animate Auditory stimuli in the AR group (p= 0.057). Additional analyses have explored differences between the AR and LR groups on non-specific measures of stress response and broad neurobehavioral status on the NNNS.

Conclusions:  This preliminary study suggests that infants at risk for autism show a diminished response to animate but not to inanimate stimuli in the newborn period. Other research has examined social responses in early to middle infancy in at risk babies, but results have been inconsistent and often subtle. While these findings require replication, they suggest that the nature of observational methods (i.e., the design of the behavioral assay) will affect the sensitivity to detect differences associated with risk for autism in very early infancy.