Associations between Sleep Structure and EEG with Daytime Functioning in ASD

Background: Sleep significantly interacts with daytime functioning. This will be illustrated by a series of studies in school-age children and young adults with ASD by describing the associations between, on one hand, sleep structure and EEG and, on the other hand, clinical measures, cognitive performance, and sensori-motor performance.

Objectives: The objectives of these studies were to elucidate the relationships between sleep structure and daytime functioning using gold stand sleep assessment methods.

Methods: Groups of 11-14 ASD children (mean age 10.3, SD = 2.2, range: 6–13) and 15-18 ASD adults (mean age 21.7, SD = 3.5, range: 16-27) were diagnosed with AID-R and ADOS. To identify relationships specific to autism, we recruited participants with ASD free from comorbidities including Intellectual Disability, epilepsy, and other psychiatric diagnoses. Results were compared to 13-15 neurotypical (NT) children (mean age = 10.2, SD = 2.0, range 7–12) and 14-17 NT adults (mean age 21.7, SD = 3.5, range: 16-27). All participants were recorded in a sleep laboratory using polysomnography (PSG) for two consecutive nights; only sleep/EEG data from night 2 were used. Behavioral scales were filled during the previous 2 weeks. Neuropsychological tests (adults only) were administered in the mornings following nights 1 and 2.

IQ measures were generally comparable between ASD and NT groups. Both age groups of participants with ASD showed more internalizing problem behaviors than the NT groups. Neuropsychological performance was lower in adult participants with ASD, particularly in tasks at least partly based on motor output.

Results: Sleep structure in participants with ASD differed from that of NT participants in both age groups, with long sleep latencies, more light nonREM (N1) sleep, and low levels of deep nonREM (N3) sleep; measures of REM sleep were comparable. Light nonREM sleep correlated negatively with IQ scores (Wechsler’s and Raven’s) in both groups of children. No significant correlations were found in adults but light nonREM sleep correlated negatively with reaction times on selective attention and procedural learning tasks. Deep nonREM (N3) sleep was negatively correlated with internalizing problem behaviors in children while REM sleep showed this association in adults. Deep nonREM (N3) sleep was negatively correlated with reciprocal social interactions core (ADOS) and positively correlated with repetitive behaviors (ADI-R).

EEG sleep spindles were less frequent in children (frontal electrodes) and adults (central electrodes) with ASD when compared to NT participants. Fewer sleep spindles were associated with (1) more trials required for learning a sensori-motor procedural task (rotary pursuit) and (2) higher IQ scores. Slow (delta) EEG activity was (1) lower in adult participants with ASD, (2) uncorrelated with IQ, and (3) negatively correlated with sensori-motor performance.

Conclusions: Signs of poor sleep are associated with impaired daytime performance in ASD as well as NT groups of participants, regardless of age. Patterns of relationships, however, differ according to age and clinical status. These results illustrate the atypical development of cortical afferent and efferent systems in ASD. They also raise the possibility that these systems act differently to regulate behavior in different age and clinical groups.