Keywords
autism - EEG - extreme spindles - neurodevelopmental disorders - sleep spindles
Introduction
Sleep spindles are powerful synchronous bursts of activity in stage 2 of sleep, between
10 and 14 Hz frequency, with the maximum amplitude in the central leads. Sleep spindles
are generated from the thalamocortical relay cells in the thalamic reticular nucleus
and are observed as the surface correlates of these neuronal oscillations in the thalamus.[1]
[2]
Unlike normal sleep spindles, occasionally there can be an activity that is of high
amplitude (up to 200 microvolts) and is more widely distributed. Such sleep spindles
of high amplitude and wide distribution are called “extreme spindles.”[3]
[4] Extreme spindles are characterized by their diffuse expression (14 to 16 Hz), continuous,
fast waves with high amplitude (200–500 microvolts), and a sharp morphology, during
light sleep[3]
[4]
[5] and sometimes persistence in the awake state. These can last up to 20 seconds.[6] It has been speculated that extreme spindles are caused by the disruption of regulatory
mechanisms, including GABAergic inhibitory circuits,[6] but the exact mechanism responsible for extreme spindles is unknown. It has no association
with epilepsy.
Extreme spindles are known to be associated with neurodevelopmental disorders, predominantly
intellectual disability (ID), Costello syndrome (ID, developmental delay, unusually
flexible joints, and facial dysmorphisms), malformations of cortical development,
and children with autism spectrum disorder (ASD).[5] It is also described in neuroinfections,[7] infantile neuroaxonal dystrophy, Menke's kinky hair syndrome, congenital muscular
dystrophy, and anti-N-methyl-D-aspartate receptor encephalitis.[8]
In children with ASD without epilepsy, electroencephalogram (EEG) is not routinely
performed in clinical practice. Even if done, it is difficult to perform a sleep EEG
on these children. Hence, this interesting finding of extreme spindles is not commonly
encountered. In our center for children with ASD, we follow a standard research protocol
of awake and sleep EEG, preferably spontaneous sleep record, lasting from 1 to 4 hours.
With this protocol, we have observed this phenomenon in children with ASD and related
disorders. In this article, we present a case series of children with ASD/related
disorders with extreme spindles and explore the possible mechanisms.
Case Series
Two-hundred eighty-nine children underwent EEG at our center of which there were 163
children with ASD and 126 children with other neurological/neurodevelopmental disorders.
Of these, 6 children had extreme spindles. Their ages ranged from 2 years, 4 months
to 9 years, 4 months. Five were male children and one was a female child. Five had
a primary diagnosis of ASD and one had global developmental delay. Two children had
a history of developmental regression. All six children had cognitive delay. Two children
had sleep difficulties, predominantly insomnia. None of the six children had a history
of seizures ([Table 1]).
Table 1
Clinical profiles of children with ASD and extreme spindles
|
Case no.
|
Age/sex
|
Primary diagnosis
|
Clinical presentation
|
History of seizures
|
Medication
|
EEG finding
|
|
1
|
4 years, 3 months/male
|
GDD
|
GDD with dysmorphic features, hyperactivity, and significant speech delay
|
No
|
None
|
Moderate cerebral dysfunction (suggested by background theta activity) and bifrontal
epileptiform discharges with extreme sleep spindles
|
|
2
|
9 years, 4 months/male
|
ASD
|
ADHD and behavioral issues, with a diagnosis of ASD with significant speech delay
|
No
|
Methylphenidate 10 mg/day
|
Otherwise, normal study with extreme sleep spindles
|
|
3
|
4 years, 3 months/male
|
ASD
|
Significant speech and language delay diagnosed with ASD and obesity
|
No
|
None
|
Otherwise, normal study with extreme sleep spindles
|
|
4
|
3 years, 2 months/male
|
ASD
|
Developmental regression since 1 year, 6 months of age, significant language and cognitive
delay, with a diagnosis of ASD. Sleep disturbances (predominantly insomnia) were reported
since onset of regression
|
No
|
Ayurvedic medication
|
Otherwise, normal study with extreme sleep spindles
|
|
5
|
2 years, 4 months/female
|
ASD
|
Developmental regression, significant language and cognitive delay, autism features,
hyperactivity, and sleep disturbances (predominantly insomnia) of 6 months duration
|
No
|
None
|
Background asymmetry with extreme sleep spindles
|
|
6
|
7 years, 5 months/male
|
ASD
|
Poor response to therapy, significant speech, and language delay
|
No
|
None
|
Otherwise, normal study with extreme sleep spindles
|
Abbreviations: ADHD, attention deficit hyperactivity disorder; ASD, autism spectrum
disorder; GDD, global developmental delay.
On examination of our case series, three children had type 3 extreme spindles, two
had type 6, and one had type 2b with a lower amplitude as per the classification by
Gibbs and Gibbs ([Figs. 1]
[6]
[Table 2]).[3]
[Figs. 1]
[6] are displayed in bipolar, longitudinal montage, at a sensitivity of 7.5 μV/mm and
filter settings of 0.5-70 Hz. Each image is from each patient in the case series ([Table 1]) demonstrating the type of extreme spindle as per Gibbs and Gibbs.[3] In our small series, type 3 spindles, which are of high amplitude and spiky appearance
resembling epileptiform discharges, were the most common type and were seen in three
children. Most of these spindles were seen in stages 2 of nonrapid eye movement sleep.
All children with type 3 and 6 had spindles with an amplitude ranging between 200
and 350 microvolts. One child also had bifrontal epileptiform discharges in the absence
of clinical seizures. Another child had asymmetric background activity. In the remaining
four children, extreme spindle activity was the only abnormality in the EEG.
Fig. 1 Type 2b extreme spindles with lower amplitude.
Fig. 2 Type 3 extreme spindles.
Fig. 3 Type 3 extreme spindles.
Fig. 4 Type 3 extreme spindles.
Fig. 5 Type 6 extreme spindles.
Fig. 6 Type 6 extreme spindles.
Table 2
Types of extreme spindles as described by Gibbs and Gibbs3
|
Type
|
Description
|
|
1
|
12–14 Hz, almost continuous, high voltage spindles in sleep
|
|
2
|
Similar to type 1 but particularly fast activity in drowsiness and sleep (further
divided as 2a and 2b)
|
|
3
|
“Spiky” high voltage spindles that resemble epileptiform discharges
|
|
4
|
Spindles of high voltage that are fast in early sleep stages but slow down to 6Hz
in deeper stages (further divided as 4a, 4b, and 4c)
|
|
5
|
Very frequent, almost continuous slow spindles of 5–7 Hz frequency
|
|
6
|
Mixture of two or more types
|
Discussion
This is probably the first of such series reported from India on the ASD population.
In our study sample, of the six children with extreme spindles, four of them did not
have any other abnormalities on EEG. One had background asymmetry and another had
background slowing with epileptiform discharges. Five of six children were males and
three of six children had hyperactivity as comorbidity. One child was on medication.
As described by Gibbs and Gibbs, we found three different types of extreme spindles
in our case series (types 2b, 3, and 6). We did not find extreme spindles in any of
the other children who underwent EEG at our center with a primary diagnosis other
than ASD including cerebral palsy, attention deficit hyperactivity disorder, learning
disability, epilepsy, psychosis, tics, or pseudoseizures.
The significance of the presence of sleep spindles and their clinical implications
in children with ASD are not clearly known. Extreme spindles have been previously
reported in children with ASD, mental retardation, and other neurodevelopmental disorders.
Children with IDs are known to be associated with a wide variety of genetic and developmental
alterations. The presence of significant cognitive delay resulting in comorbid ID,
especially in children with regression and severe forms of ASD with multiple comorbidities,
could be a contributory factor to the presence of extreme spindles.[5]
[9] Some studies have also mentioned the presence of a lower density of sleep spindles
and fewer spindles over the central and prefrontal areas.[1] Too few spindles or extreme spindles are also both associated with ID.[1]
[5] In this case series, all children with ASD had ID as comorbidity. This could be
one of the factors that are contributing to the occurrence of extreme spindles rather
than fewer spindles or spindles with lower density.
Extreme spindles must also be differentiated from spindle coma, beta coma, and alpha
coma ([Figs. 7], [8]). Spindle coma is different from extreme spindles with characteristics of 11 to
14 Hz spindle discharges in comatose patients where the background is predominantly
theta or delta. EEG patterns have frequent bursts of sleep-like activity, mostly due
to the presence of thalamocortical circuits and raphe nuclei activity, but the absence
of activity in the reticular activating pathways in the midbrain.[10] Alpha coma, on the other hand, is an EEG pattern seen in cases of profound coma
as a result of trauma, encephalitis, drug overdose, or hypoxic-ischemic encephalopathy
that is seen as a diffuse band of alpha frequency across all the leads.[11] Extreme spindles must also be differentiated from beta coma.[12] This occurs typically due to benzodiazepine toxicity or barbiturate use and is seen
as discharges of 13 to 30 Hz with low amplitude beta activity that overrides normal
activity throughout the recording ([Fig. 8]).
Fig. 7 Example of alpha coma.
Fig. 8 Example of beta coma.
In conclusion, children with ASD can have multiple overlapping neurodevelopmental
comorbidities, sleep disorders, and epilepsy co-occurring with the core symptoms.
The shared biological pathways may contribute to the occurrence of these extreme spindles.
There is a need to further understand these findings from clinical, genetic, and neurological
perspectives to inform clinical care, and interventions.
