Although epilepsy has been known since the 1940’s to be associated with autism, it is seen only in 10-30% of cases. Motor seizures are less frequent than the presence of nocturnal epileptiform EEG abnormalities in sleep. There has been no clinical research correlating these EEG changes with sleep disruption in the absence of clinical seizures. There are increase of sleep disruption in patients with known epilepsy, but it is unclear if this is due to interictal (between attacks) epileptiform activity of the brain, or possible effect of medications, mood swings, or other factors. In autism, these factors of EEG and how it may affect sleep activity is unknown.
Autistic patients report waking up between 2-5 AM, sometimes from night terrors, and/or being afraid to go back to sleep. The patients were not reported frequently of having motor seizures. Often these night patterns related to the deterioration in autistic behavior in these children. The only way to really diagnose these patterns on EEG accurately in my opinion are to get a multichannel overnight sleep EEG with at least three REM cycles of sleep.
The most complete studies of prolonged overnight EEG of autistic children have shown that 30 to 61% may have epileptiform EEG activity in sleep in the absence of motor seizures. There was improvement in the most of those children who treat their EEG abnormalities with medications. Therefore, it seems prudent in autistic patients with sleep disruptions to consider an overnight EEG test with multiple channels. This is more than a typical polysomnogram offers. Certainly a polysomnogram can still be done simultaneously in cases of suspected restless legs or apnea. In fact I have often combined both procedures where appropriate.
Once the EEG patterns have been diagnosed, then treatment options could include anticonvulsants. My favorite choice is valproic acid which tends to change the EEG. There have been no formal studies on the other anti-convulsants however to see if they would be any worse or better at promoting sleep or changing EEG patterns in this autistic population.
As a conclusion: neurological consultation should be sought with EEG in autistic children with sleep disruptions. If the EEG is abnormal treatment could be indicated. Many clinicians are still unaware of the importance of sleep disruptions in the autistic population.
In future columns I will discuss other sleep problems in the autistic population and treatment options as well. These will include options for sleep onset issues, causes of sleep mid-cycle disruption and insomnia, sleep apnea, restless legs, and other.