Sleep Patterns in Children with High Functioning Autism: Polysomnography, Questionnaires and Diaries in a Non-Complaining Sample

Background:  Sleep patterns in children with autism are frequently only reported on questionnaires filled by parents.

Objectives:  The aim of this study was to characterize sleep using objective and subjective measures in the same sample of autistic children without comorbidities and no primary complaints of sleep disorders.

Methods: Eleven autistic children (10.3 ± 2.2 years) diagnosed with ADI-R and ADOS criteria and 13 typically-developing children (10.2 ± 2.0 years) were recruited. Exclusion criteria comprised the use of psychotropic medication, a full IQ lower than 75, a history of epilepsy and spontaneous complaints of sleep disorders from the parents. After their children were recruited, parents filled a sleep diary for 2 weeks and the Child Sleep Habit Questionnaire (CSHQ); children were then recorded for 2 consecutive nights in a sleep laboratory. Sleep stages of night 2 were scored according to Rechtschaffen and Kales (1968) using 20 sec. epochs. Stage 2 sleep spindles and K-complexes were visually scored on bilateral prefrontal and central electrodes; REM sleep rapid eye movements were also scored. Variables were log transformed when abnormally distributed. Groups were compared using t-tests for independent samples.

Results: The CSHQ showed no significant group differences. Sleep diaries showed a longer sleep latency in autistic children (43.6±39.0 minutes vs. 17.2±17.5 minutes, p=.037). Polysomnographically recorded sleep latency was longer in autistic children (33.0±8.3 minutes vs. 14.4±4.6 minutes, p=.023), the duration of stages 3 and 4 (slow-wave sleep: SWS) was shorter (18.2±3.2 % vs 23.6±5.7 %; p=.009) but total sleep time was similar (539.7±54.3 vs. 560.8±60.2; p=.94). Sleep spindle density (per hour of Stage 2) was similar in both groups at central electrodes and Fp1 but it was lower at Fp2 (119.2±97.7 vs. 225.5±122.2, p=.03). The density of K-complexes was lower at the four electrodes (.0001<p< .01). REM sleep parameters (latency, duration, distribution, eye movement density) were not different between groups.

Conclusions: These results show that autistic children without subjectively reported sleep difficulties according to the CSHQ show signs of poor sleep on sleep diaries and polysomnography. Sleep diaries disclosed longer sleep latencies. Most polysomnographic differences were in nonREM sleep, raising the hypothesis of a difficulty for autistics to synchronize their cortical activity, possibly leading to impaired cortical sleep protective mechanisms such as low densities of sleep spindles and K-complexes.