Electrophysiological Assessment of Low-Contrast Visual Function and Neural Noise in Children with Autism Spectrum Disorder

Background:  Hyper- and hypo-reactivity to sensory input is now included in the DSM-5 as a core deficit of autism spectrum disorder (ASD). Previous studies, based on parent-report questionnaires, indicate that sensory symptoms are significantly more prevalent in children with ASD (60-90%) as compared to children with other developmental delays and typically developing controls. Furthermore, younger children who are more severely affected have the greatest number of sensory symptoms. Electroencephalograms (EEG) offer an objective method to assess sensory processing; however, lower-functioning populations are often difficult to test using conventional EEG methods, which usually involve long-duration runs, many electrodes applied to the head, and behavioral tasks. The current study explores the feasibility of single-channel EEG recording to investigate neural mechanisms and pathways associated with low-contrast vision and neural noise in children with ASD.

Objectives:  (1) To assess the integrity of low-contrast responses in ON- and OFF-cell pathways (presumably within the magnocellular system) and (2) to assess the level of neural noise within the visual cortex of children with ASD.

Methods: Visual evoked potentials (VEPs) were obtained from children with ASD and typically developing controls by extracting the VEP from the ongoing EEG using a single electrophysiological channel comprised of three electrodes placed on the midline of the scalp. Stimulus conditions included contrast sweeps of bright or dark isolated-checks with appearance/disappearance sinusoidal modulation (12.5 Hz) to examine ON and OFF pathways, respectively. Contrast increased in 1-s octave steps from low to high levels. Each swept-parameter condition consisted of ten 7-s runs. ASD diagnoses were determined based on standardized research diagnostic instruments (ADOS, ADI-R) and DSM-5 criteria.

Results:  Children in the ASD group displayed deficits in low-contrast responses to both bright- and dark-check stimuli. These deficits were present across contrast levels (1,2,4,8,16,32%) for both conditions. Results also indicate that the ASD group had higher levels of neural noise than did the control group at low contrasts (< 8%) when significant responses (signal-to-noise ratios >1) were not obtained. The ASD group displayed a relatively constant level of noise across the range of contrasts used, whereas the control group displayed low levels of noise at low contrasts and increases in noise level at increasing contrasts. Thus, group differences in levels of neural noise were significant only under low contrast conditions in which responses were either absent (mean signals not above noise) or weak.

Conclusions:  Results provide electrophysiological evidence for deficits in low-contrast (magnocellular-biased) visual function and increased neural noise in children with ASD. This study also demonstrates the feasibility of short-duration, single-channel VEP recording in children at various levels of cognitive and behavioral functioning. VEPs hold promise as a rapid and reliable biomarker of early-stage visual processing deficits in ASD. Future studies must assess whether these findings persist in larger samples and whether abnormalities in basic low-level/non-social stimuli may translate into higher-order perceptual differences.