Clinical manifestations

Presentation and course

Patients with epilepsy may present with three major complaints about sleep: excessive daytime sleepiness, difficulty initiating or maintaining sleep, or unusual nocturnal events. These complaints are seen in both children and adults and deserve evaluation. Patients with epilepsy may have difficulty perceiving or conveying their sleep-related symptoms. Yet, those who have sleep complaints are noted to have poorer quality of life, more complaints, and depression (22).

Sleep complaints are common among patients with epilepsy, yet patients and clinicians may falsely assume that the sleep-related symptoms are a common part of their epilepsy (06; 48). Clinicians should be astute to less obvious manifestations of sleep issues, especially when seizure frequency increases. Patients may exhibit an increase in seizures related to sleep disorders, sleep deprivation, or sleep disruption. Brief questioning of the patient or caretaker regarding the patient’s satisfaction with sleep, daytime sleepiness, or the presence of snoring may direct the clinician to potential sleep issues. Sleep issues are common and may also present differently in children. Children may have the classical symptoms of sleep complaints of shorter sleep duration, excessive sleepiness, nightmares, and sleep disordered breathing (43). However, children may not be able to verbalize that they are sleepy or have trouble sleeping, but they may show resistance to awakening in the morning, greater daytime irritability, poor behavior, decreased school performance, or hyperactivity. Some of these sleep issues appear present prior to the onset of the epilepsy. Vatansever Pinar and colleagues found that children with new onset epilepsy were more likely to have sleep issues, and these issues tended to improve with treatment of the epilepsy (60). Investigators found that children with ongoing epilepsy had more sleep problems, and it was postulated that epilepsy caused further sleep disruption (64). In addition, sleep activated epilepsies appear to disrupt circuits and predict cognitive deficits in the daytime (29).

Excessive daytime sleepiness is a common complaint, ranging from 24% to 50% of patients with epilepsy (19). Patients may relate falling asleep during inappropriate situations. Caregivers may be able to help identify the patient’s unwarranted ability to fall asleep. Patients who are sleepy should be asked about total time devoted to sleep, their sleep schedule, relationship to medication doses, symptoms of snoring, or excessive movement during sleep. Patients with frequent seizures may develop daytime sleepiness related to their seizure activity. Scales such as the Epworth Sleepiness Scale provide a measure to track the subjective sleepiness). Similarly, tools such as sleep diaries can provide insight into the causes and relationships of schedule to sleep and seizures. In a review of sleep and seizure diaries examining seizure frequency compared to sleep, Cobabe and colleagues found that sleep diaries provided important information regarding altered sleep times and frequent napping as clues contributing to lower quality of life and seizure control (11). Scales such as the Epworth Sleepiness Scale may provide a measure to track the subjective sleepiness. This sleep disturbance also extends into other family members. Hamamci and colleagues found that similar to patients with epilepsy, spouses of patients with epilepsy have poorer quality of sleep, greater complaints of sleep disturbance, and higher sleep medication use than spouses of patients without epilepsy (24). These issues were even more prominent in those with spouses having recurrent seizures. Yet, studies measuring objective sleepiness are limited.

One-third of the general population and over 50% of patients with epilepsy complain of insomnia (50). Insomnia is characterized by difficulty initiating or maintaining sleep, with the complaint of feeling unrefreshed or having daytime sequelae such as decreased daytime performance, sleepiness, or fatigue. This complaint may be a reflection of underlying sleep disorder, the epilepsy, or the medication and is more common in those with intractable epilepsy (50). Insomnia and poor sleep are frequently associated with poor seizure control and lower quality of life (48). Patients describe difficulty falling asleep and frequent arousals at night. These symptoms necessitate careful review of daily habits, activity, diet, caffeine intake, timing of medications, supplements, beliefs regarding sleep, the sleep environment, and bedtime rituals undertaken when sleep is poor. Similarly, depression and anxiety increase the risk of insomnia (27). These mood issues are common in patients with intractable seizures.

Patients should be questioned regarding their nighttime sleep. Bed partners may note snoring, pauses in breathing (apneas), or excessive movement. These may be good clues to utilize the STOPBANG questionnaire or the Sleep Apnea Scale of the Sleep Disorders Questionnaire, which has proven to be an effective tool for determining the risk of sleep apnea in this population (52). Additional clues may include morning headache or unrefreshing sleep. In children, hyperextension of the neck and head during sleep may be a subtle sign of obstruction.

Patients with epilepsy have more movements in sleep than control patients (17). Some of these movements may be brief seizures or related to the frequent arousals. Nocturnal events are also common in patients with epilepsy. These events may be recurrent seizures, but they may also represent other parasomnias. Patients with nocturnal seizures may have events that are similar to their daytime events or may encompass different behaviors. Seizures may be relatively subtle, such as lip smacking or blinking, or they may be more complex behaviors, including posturing, rocking, or vocalizations. Sleep-related hypermotor seizures, previously known as nocturnal frontal lobe epilepsy, are a classic example of the variety of these events (33). However, the key feature is that the behaviors are stereotypic in that the behavior or some component of the behavior is the same with each event. In addition, some patients may only note the aftermath of a seizure, such as urinary incontinence, tongue sores, or feeling sore and mentally foggy in the morning.

Sleep-related epilepsy syndromes can demonstrate an inheritance pattern. A subset of sleep-related hypermotor seizures, autosomal dominant nocturnal frontal lobe epilepsy, presents in a variety of motoric events and sudden arousals that may be confused with a parasomnia (33). Similarly, benign childhood epilepsy with centrotemporal spikes (formerly Rolandic epilepsy) has familial pattern and occurs in younger school-aged children and is associated with nocturnal seizures that manifest as hemifacial spasms, drooling, and speech arrest. These events typically occur about 20 to 90 minutes after bedtime and are responsive to medication. Panayiotopoulos syndrome presents most commonly in children as nocturnal paroxysmal autonomic symptoms, such as episodic nocturnal vomiting.

Seizures may also be influenced by the time of day. Circadian and ultradian patterns in seizures and interictal discharges appear to depend on the location of the seizure focus (53). Examining circadian preference, Kendis found that patients with generalized epilepsies were more likely to be night owls than patients with focal-onset epilepsy (28). Although there may be some relationship to the awakening nature of the generalized epilepsy, the linkage is unclear.

Table 1. Epilepsy Syndromes Associated with Sleep

Epilepsy syndrome	Association to sleep/EEG findings	Characteristic features
Focal onset
Sleep related hypermotor epilepsy (SHE) or autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE)	Occurs predominantly during sleep	-Paroxysmal arousals with recurrent motor behavior -Episodic stereotyped nocturnal wandering -Paroxysmal arousals with dystonic/dyskinetic postures
Benign childhood epilepsy with centrotemporal spikes (BECTS)	70%-80% of seizures exclusively during sleep, centrotemporal spikes with transverse dipole	-Clonic movements of mouth with drooling -Hemifacial and tonic body activity -Speech impairment
Childhood occipital lobe epilepsy (Panayiotopoulos type)	Nocturnal seizures shortly after falling asleep, occipital spikes	-Nocturnal vomiting with gazing to one side with rhythmical muscle contractions
Generalized
Juvenile myoclonic epilepsy (JME)	3.5-4.5 Hz generalized spike and wave discharges	-Single and multiple myoclonic jerks typically 1-2 hours after awakening
Undetermined
Epilepsy with continuous spike and wave during sleep (ESES)	Continuous spike and wave discharges during sleep	-Nocturnal focal motor seizures -Cognitive decline

Prognosis and complications

The prognosis of patients with epilepsy and sleep complaints is highly dependent on the underlying sleep issue. Sleep disruption such as insomnia is associated with more intractable epilepsy (48). Some sleep issues such as medication-related effects are easily handled with change of timing or dose of medication. Touchon and colleagues found that treatment of epilepsy may improve sleep (59). Other issues such as intractable epilepsy inducing sleep fragmentation may not be as easily treated, yet surgery for temporal lobe epilepsy may improve sleep architecture and total sleep time (67). Additionally, treatment of sleep disorders such as sleep apnea can add long-term benefits for reducing seizures and improving sleep (49). For most sleep complaints there are therapies that improve the quality of life for patients without epilepsy, and this would appear to give hope to improving the quality of life for patients with epilepsy. Newer antiseizure medications appear to have fewer side effects of sleepiness and may improve sleep architecture (34). Patients with epilepsy, however, frequently develop poor sleeping habits and need to be reminded of the importance of good sleep hygiene (12). With persistent management and questioning, most patients notice improvement in their symptoms.

Newer work has implicated that epileptic discharges during sleep may have an influence on daytime cognitive performance. As sleep is involved in neural regulatory functions in preparing neuronal circuitry for optimal performance during wakefulness, the interruption of these processes has been implicated in lower cognitive function (23). This is most notable in the children with benign childhood epilepsy with centrotemporal spikes, in that more active interictal activity is associated with cognitive dysfunction. Similarly, the effect of treatment appears to improve daytime cognitive function (42). Similar findings have linked interictal activity at night and associated memory issues (32).

Clinical vignette

A 23-year-old right-handed man with myoclonic epilepsy presented with an increase in his morning myoclonic jerks and noted the return of his generalized tonic-clonic seizures. His seizures were previously controlled with valproate 500 mg, three times a day. He also noted that his sleep was unrefreshing, and he had tripled his typical coffee intake to stay awake during work. He had gained approximately 30 pounds over the last two years, and his family complained of his loud snoring. Polysomnography showed an apnea-hypopnea index (number of apneas and hypopneas per hour) of 25.6 events per hour, with a minimum oxygen saturation of 85%. A trial of continuous positive airway pressure at 10 cm corrected the respiratory disturbance and snoring.

Once on continuous positive airway pressure therapy, the patient noted an immediate reduction in his myoclonic jerks of greater than 90% and no further generalized seizures. The patient also noted a return of feeling refreshed in the morning, and he was completely off caffeine. His valproate levels remained unchanged.

Biological basis

Etiology and pathogenesis

	• Epileptic discharges are more common in NREM sleep and less common in REM sleep.
	• Epileptic discharges in REM sleep are more likely to localize close to the epileptic focus.
	• Epileptic discharges are associated with great fragmentation of sleep.

Patients with epilepsy have more disturbed sleep. Touchon and colleagues noted that sleep fragmentation is more common in individuals with frequent seizures and that the fragmentation is improved with therapy and seizure control (59). This disturbance appears related to both ictal and interictal discharges. Seizures, although causing postictal somnolence, increase wake time and arousals from sleep and decrease REM sleep (41). Similarly, interictal discharges appear to cause more arousals and sleep fragmentation. This sleep fragmentation is seen clinically and is extremely similar to that seen in a primate model of diurnal cortical seizures (63). In this model, the cortical induced seizures appeared to affect the circuits involved in REM sleep entry and also another circuit controlling nocturnal waking. These regulatory changes are most likely related to nonsense signals from the epileptic discharge disrupting the typical physiological networks of sleep and wake. This disruption may be more common if the epileptic focus has richer connections with the hypothalamus (55).

Sleep disorders may also be prominent in patients with epilepsy. Several investigators have found that patients with epilepsy had a higher than expected rate of obstructive sleep apnea and movements during sleep, and these findings appear to be more common in those with intractable epilepsy (36). Yet patients with epilepsy do not appear to have a higher than expected prevalence of central apnea (62). Although some reports have suggested a link between sleep-related breathing disorders and sudden unexplained death in epilepsy, these are primarily retrospective studies using a SUDEP inventory tool that includes many of the risk factors for sleep apnea (10). One prospective study examining this relationship was unable to demonstrate a link of sleep-related breathing issues and the risk of sudden unexplained death in epilepsy (07), and another cross sectional study of 95 participants failed to show an association of AHI to sudden unexplained death in epilepsy risk assessment (47).

Sleep has an effect on both interictal and ictal discharges. NREM sleep appears to be more permissive of seizure onset. Several studies have shown that of all the NREM stages, particularly stage 2 sleep, have greater propensity for seizure occurrence compared to wakefulness. REM sleep, being the least likely state for seizures, appears to inhibit seizures (20). REM sleep deprivation in rats increases the kindling process, and increasing REM sleep appears to slow the kindling process (30). Similarly, NREM sleep increases interictal discharges and increases the number of locations these discharges are seen. This may be related to an increase in neurons available for recruitment into the discharge or thalamocortical network changes. REM sleep, however, decreases the frequency and number of locations for interictal discharges and can be valuable in localizing the epileptic focus (39). The influence of sleep on epilepsy is supported by observations that in specific epileptic syndromes, seizures occur exclusively or primarily during NREM sleep. Neuronal synchronization within thalamocortical networks during NREM sleep results in enhanced neuronal excitability, leading to more diffuse distribution of focal discharges as well as facilitation of seizures and interictal epileptiform discharges in many persons with partial epilepsy. This feature has usefulness in identifying the seizure focus. Work by several investigators showed that focal-onset interictal discharges in REM sleep are highly likely to occur at the seizure focus (51). However, sleep deprivation appears to increase cortical excitability in patients with idiopathic generalized epilepsy (03), although most seizures occurring in sleep appear to initiate from the frontal and temporal lobes (25). Similarly, treatment of obstructive sleep apnea with CPAP appears to reduce interictal activity (49).

Epilepsy also influences sleep. Several investigators have shown that seizures cause more frequent arousals and awakenings. Nobili and colleagues showed that seizures cause REM sleep suppression, even if the seizure occurs during the day prior to sleep (41). In addition, patients may demonstrate paroxysmal arousals during sleep associated with abnormal motor activity, some of whom had ictal or interictal epileptiform abnormalities. Genetic analysis of kindred with autosomal dominant form of sleep related hypermotor epilepsy has revealed some exciting clues as to the links of some nocturnal epilepsies. This syndrome appears to be linked to several genes that produce a change in the nicotinic receptor, rendering it hypersensitive to acetylcholine (05). The hypersensitivity of the acetylcholine receptor appears to be manifest as seizures during or following brief arousals.

Investigators have shown that daytime generalized seizures globally increased slow wave activity during sleep and that focal interictal spikes in the last hour of wakefulness increased slow wave activity localized to their seizure focus (08). Similarly, interictal spikes during NREM sleep may interrupt the homeostatic drive of slow wave sleep and reduced daytime learning. This hypothesis has been challenged by work from Eriksson who was unable to show differences in 19 children with focal onset epilepsy compared to controls in slow wave activity in the first hour of NREM sleep (14). Also, they were unable to show any connection in IQ to measures of slow wave activity in patients or controls. Similarly, Fonseca and colleagues found that sleep quality did not independently impair cognitive function in those over the age of 12 years (15). This link between sleep disruption and cognition may be dependent upon the type of epilepsy and the circuit involved. Lambert found that memory issues were associated with frequent interictal discharges in sleep in individuals undergoing intracranial monitoring for epilepsy surgery evaluation (31). Thus, further work is needed to elucidate if both seizures and interictal discharges can influence the homeostatic drive of sleep and have negative impact on daytime function.

Epilepsy may influence the circadian rhythm also in a bidirectional manner (04). Some forms of epilepsy are more prominent at specific times in the circadian rhythm owing to the influence on the circadian rhythm on the underlying type of epilepsy. The corollary also exists in animal studies, suggesting that frequent interictal discharges and seizure events disrupt the circadian regulation as well as the output of the endocrine and autonomic nervous system. The circadian disruption also appears to play a role in epileptogenesis. Animal models with loss of specific circadian proteins develop seizures in vivo. Overall, this complex relationship deserves more research to identify potential therapeutic opportunities.

Epidemiology

	• Sleep issues are more common in patients with epilepsy.
	• Patients with intractable epilepsy have greater frequency of sleep complaints, which impacts their quality of life.

National Health Interview Surveys performed in the United States in 2013, 2015, and 2017 showed that adults with epilepsy were more likely to report short or long sleep and poorer sleep quality (57). Sleep issues appear greater in adults and children and adults with epilepsy, and the issues appear across the spectrum of sleep complaints. Both children and adult patients with intractable epilepsy appear to have a greater number of sleep complaints (06; 16; 64). Bergmann in adults and separately Winsor and Furones Garcia in children showed that patients with epilepsy are more likely to have sleep disorders and daytime sleepiness, impacting daytime function. Similarly, sleep disturbance was more than twice as prevalent in persons with epilepsy compared to controls. Others have shown that patients with partial epilepsy have significantly more symptoms of insomnia, periodic leg movements, excessive daytime sleepiness, narcolepsy, apnea, and snoring, although these symptoms were not confirmed by polysomnography. No matter the cause of shorter total sleep time, poor sleep efficiency and prolonged sleep latencies are associated with impaired executive function and memory issues in patients with refractory epilepsy (09).

Prevention

• Patients with epilepsy should be advised on keeping consistent sleep schedules, sleep hygiene, and optimization of medications.

Prevention of sleep complaints in patients with epilepsy should be considered in three areas: consistent sleep schedule, sleep hygiene, and optimization of medications. Patients who are not working may not keep a routine schedule of sleep and wake, nor dedicate enough time to sleep. Consistent sleep-wake schedule across the 7-day week can help the circadian rhythm. Patients should be advised to avoid sleep deprivation and oversleeping because both may increase the likelihood of seizures. Stirling and colleagues, using data from wearable devices, found that sleep duration had less of an effect on seizure frequency, whereas inconsistent bed times and wake times were associated with higher risk of seizure recurrence (54). Practicing good sleep habits such as minimization of caffeine intake and stimulating activities before bed, inclusion of exercise during the day, and avoiding napping will ensure patients have more restful sleep (20). Patients should also be advised to avoid developing obesity because this may increase the risk of sleep apnea and hypoventilation. If possible, sedating medications should be dosed prior to sleep, and alerting medication should be dosed early in the wake period. These measures will help reduce the likelihood of side effects impairing sleep or wake. Controlling the seizure activity will add to improving sleep. Patients should also have somnogenic medications maximized in the nighttime and alerting agents during the day to improve the sleep-wake cycle.

Differential diagnosis

The differential diagnosis is somewhat different for each specific sleep complaint. For hypersomnia in patients with epilepsy, the clinician should consider the common contributing features: total amount of sleep, disturbance of sleep, circadian rhythm disorder, dysfunction of the brain awake, or sleep mechanisms. Patients should be questioned as to whether sleep is refreshing, indicating the duration of sleep and whether quality of sleep is adequate. Patients should also be questioned regarding symptoms of other sleep disorders such as sleep apnea or restless legs syndrome, sleep schedule changes, medications, and features of narcolepsy. Similarly, for symptoms of insomnia, patients should be questioned regarding features of sleep habits, beliefs, schedule, medication use, psychiatric symptoms, and other medical changes. Many patients with insomnia develop maladaptive behaviors that interfere with sleep, such as napping, having the television on during sleep, or sleeping in locations other than the bedroom. Patients with epilepsy may have excessive caffeine use or incorporate other supplements that may be stimulating. Additionally, timing of medication may be important. Medications that cause alertness should be maximized during the wake period, and somnogenic medications should be maximized prior to sleep onset. Nocturnal events may include recurrent seizures, NREM and REM parasomnias, psychiatric events, or medical conditions. Disorders of arousal can be provoked by epileptic activity, and some rare seizures may present with symptoms consistent with REM sleep behavior disorder (13; 02). A detailed history is key to determining the need for further evaluation.

Table 2. Key Features of Nocturnal Events

Sleepwalking Symptoms: Slow, deliberate, complex behaviors Time of night: First half of sleep period Duration: Seconds to minutes Frequency: Less than one per night to fewer Stereotypic: No Memory: No or partial vague memory Eyes: Open Polysomnographic findings: Arousal from slow-wave sleep

Sleep terrors Symptoms: Piercing scream, followed by fight or flight response Time of night: First half of sleep period Duration: Seconds to minutes Frequency: Less than one per night or fewer Stereotypic: No Memory: No or partial vague memory Eyes: Open Polysomnographic findings: Arousal from slow-wave sleep

Confusional arousals Symptoms: Variety of unusual behaviors on sudden awakening Time of night: Anytime Duration: Seconds to minutes Frequency: Less than one per night or fewer Stereotypic: No Memory: No or partial vague memory Eyes: Open Polysomnographic findings: Arousal from slow-wave sleep

Sleep-related eating disorder Symptoms: Eating of high-calorie or strange foods in a messy manner Time of night: First half of sleep period Duration: Minutes Frequency: May occur nightly Stereotypic: No Memory: No or partial vague memory Eyes: Open Polysomnographic findings: Arousal from NREM sleep

Sleep enuresis Symptoms: Uncontrolled urination in sleep Time of night: Anytime Duration: Minutes Frequency: May occur nightly Stereotypic: N/A Memory: No or partial vague memory Eyes: N/A Polysomnographic findings: Occurs more commonly from NREM sleep, but can occur in REM sleep

REM sleep behavior disorder Symptoms: Dream enactment, may be violent Time of night: Latter half of sleep period Duration: Seconds Frequency: Nightly or multiple times per night Stereotypic: No Memory: Yes Eyes: No Polysomnographic findings: Loss of REM sleep atonia

Nightmares Symptoms: Frightening dreams associated with anxiety Time of night: Latter half of sleep period Duration: Seconds to minutes Frequency: Variable Stereotypic: No, but may have a common theme Memory: Yes Eyes: No Polysomnographic findings: Occur in REM sleep

Seizures Symptoms: Variable Time of night: Variable Duration: Seconds to minutes Frequency: Variable Stereotypic: Yes Memory: Usually no Eyes: Varies Polysomnographic findings: Recorded events are behaviorally consistent with description

Diagnostic workup

	• Patients with epilepsy should be questioned about sleep symptoms.
	• If sleep symptoms are raised, standard workup for sleep disorders should be pursued.
	• Patients with unexplained increase in seizure frequency should be asked about sleep issues.

The diagnostic workup should always start with a detailed history and physical examination. For hypersomnia, the historical features of total time devoted to sleep, weekday versus weekend schedule, medication use and timing, and presence of snoring are all helpful. For those with snoring or excessive movement in sleep, polysomnography should be considered. Patients with unexplained sleepiness may require an overnight polysomnogram and multiple sleep latency tests to evaluate for narcolepsy or other central causes of hypersomnia. A urine drug screen should be included. Patients may also have difficulty distinguishing fatigue from sleepiness and, thus, other endocrinopathies such as thyroid disorders and diabetes should be evaluated.

For patients with insomnia, a sleep diary may prove helpful following a detailed history of daytime and nighttime activities. Some patients may have an inactive lifestyle and nap frequently during the day, which makes sleep at night less likely to occur. Actigraphy may also help delineate the patient’s sleep schedule over a 1- to 3-week period and give clues as to the circadian rhythm. Some patients may have maladaptive behaviors such as keeping the television on during sleep or waking to eat in the middle of the sleep period. Patients also should be questioned regarding caffeine and other stimulant use. Timing of alerting or somnogenic anticonvulsants may demonstrate clues to the patient’s sleep difficulties. Patients with insomnia may indicate snoring, which would trigger the necessity of a polysomnogram to evaluate for sleep apnea. Depression and anxiety may also present as insomnia and, thus, further screening for psychiatric issues may prove helpful.

For nocturnal events, the key starting place is a clear description of the events from a witness (40). The time of the event, length of the event, memory for the event, and stereotypic behavior all give clues to the etiology. Individuals with atypical presentations, those at risk for hurting themselves or others, or those with symptoms of other sleep disorders or who report posturing, stereotypic, or very frequent events should undergo an overnight polysomnogram with extended EEG montage. Clinicians should be aware of the limitations. Although scalp EEG recording is helpful, it may not give the full picture of electrographic discharges, and some discharges may not be evident due to the limited number of participating neurons, limited field strength, or if the direction of the electrical vector is not in the field of recording. Some of the epileptic-related arousals and sleep disruptions may not be detected by scalp EEG. Video-EEG polysomnographs should have complete EEG coverage of the frontal and temporal regions and be recorded so as to allow review of EEG at the conventional paper speed of 30 mm/second. Time-locked video is key for reviewing unexplained nocturnal movements and behaviors, especially those that resemble sleep terrors or somnambulism, and in looking for epileptic characteristics and discharges. A standard diagnostic pathway for diagnosing sleep related epilepsies was proposed jointly by several European societies (40).

Management

	• Improved management of the seizures may help reduce epilepsy-related sleep difficulties.
	• Patients with epilepsy respond to standard therapies for sleep disorders.
	• Preliminary studies suggest treatment of sleep disorders appear to help patients with epilepsy.

Management of seizures is a cornerstone in helping patients with epilepsy-related sleep difficulties. The SANAD study suggests lamotrigine and carbamazepine as the best choice medications for focal-onset seizures (37; 38). Valproate and topiramate appear to be the most effective in primary generalized epilepsy, with the notation that women of childbearing potential may consider lamotrigine for the lower fetal malformation risk. With regard to sleep changes, antiseizure medication may decrease sleep fragmentation in patients with epilepsy. The barbiturates and benzodiazepines are typically sedating and suppress REM sleep. Also, somnogenic, gabapentin, and pregabalin have been found to increase slow-wave sleep in healthy adults. This finding may be useful in patients with epilepsy with suppressed slow-wave sleep at baseline. Valproate and topiramate also appear to cause an increase in somnolence in some people, but little information is known about effect on sleep stage. Levetiracetam, similarly somnogenic, was found in a study to decrease sleep stage shifts and increase total sleep time and NREM sleep, and lacosamide appears to improve sleep architecture. Although we do not have a clear understanding of sleep effects for most anticonvulsants, it is recommended that somnogenic medications be predominantly dosed at night and alerting agents during the day; this should at least mitigate some of the potential side effects.

Patients with epilepsy may develop sleep disorders that influence the epilepsy. Sleep apnea is more common in patients with epilepsy and these patients may have symptoms similar to those without epilepsy (36). Patients with epilepsy who are more likely to have obstructive sleep apnea are older, male, and have higher seizure frequency. Several investigators have shown that patients may reduce seizure frequency with therapy for sleep apnea, including patients thought to be intractable (61; 36). Using this hypothesis, researchers have proposed that improvement of other sleep issues may improve epilepsy. Although no other trials to improve seizure control through treatment of a sleep disorder are available, the application of cognitive behavioral therapy for insomnia has shown benefit on sleep parameters in patients with epilepsy, yet seizure frequency needs further study (01; 44).

Another potential mechanism to improve sleep and decrease recurrent seizures is the reinforcement of the circadian rhythm (04). The circadian rhythm is one of the major drivers of the sleep-wake cycle and appears to have bidirectional ties to epilepsy. Thus, improvement in the circadian rhythm has been postulated to impact seizure frequency. Results using bright light therapy to enhance the circadian rhythm have been mixed and may represent a difference related to the location of the seizure focus (26). The circadian regulator, melatonin, has also been debated whether it helps reduce recurrent seizures. In a small pilot study of 10 patients with intractable epilepsy ages 9 to 32 years, Goldberg-Stern found a significant reduction in the frequency of diurnal seizures with 10 mg of melatonin at bedtime (18). In another randomized double blind placebo controlled crossed over study examining the effect of 3 mg melatonin in 60 patients with idiopathic generalized tonic clonic seizures, Maghbooli found that melatonin reduced the severity of the seizures and improved sleep (35). The investigators also showed a trend toward reduction in seizure frequency with melatonin. Although these preliminary studies are encouraging, larger studies are needed to definitively show the usefulness of melatonin in patients with epilepsy.

The circadian rhythm also influences pharmacokinetics and brain responsiveness to epilepsy therapies. Studying potential chronotherapy applications in patients with epilepsy, Yegnanarayan and colleagues showed a circadian effect on phenytoin level and metabolism with more subjects in the therapeutic range with higher evening dosing (65). Two additional studies suggest that some patients with epilepsy have better seizure control with two thirds or more of the medication given in the evening (66; 21). Similarly, once a temporal pattern is recognized, higher doses can be applied prior to these more seizure prone times of day. Overall, these concepts emphasize the importance of leveraging the circadian rhythm to improve seizure control.

Table 3. Somnogenic and Alerting Anticonvulsants

Perceived sleep effect	Anticonvulsant
Somnogenic	Carbamazepine Gabapentin Levetiracetam Oxcarbazepine Phenobarbital Phenytoin Pregabalin Rufinamide Topiramate Valproate Vigabatrin
Alerting	Ethosuximide Felbamate Lamotrigine Tiagabine Zonisamide

The management of sleep disorders in patients with epilepsy is similar to that of patients without epilepsy. The clinician can look for opportunities to utilize common medications such as off-label use of pregabalin or gabapentin for restless legs syndrome or somnogenic medication to help with sleep continuity. Pregabalin and gabapentin are found to increase slow-wave sleep. For sleep apnea, CPAP can be effectively used in patients with epilepsy. Caution should be exercised when using a full-face mask in individuals unable to remove the mask on their own or in those with frequent postictal vomiting. Insomnia in patients with epilepsy can improve with cognitive behavioral therapies. After identification of the etiology, nocturnal events can be treated accordingly. Anticipatory awakenings and benzodiazepines are frequently used for disorders of arousal, and clonazepam and melatonin are effective for REM sleep behavior disorder. All patients should be followed closely to ensure adequate treatment effect.

Special considerations

Pregnancy

Women with epilepsy report more difficulty with sleep and reduced quality of sleep during pregnancy and postpartum compared to healthy pregnant women (58). Although not well understood, this difference in sleep may be multifactorial. For some women, pregnancy may be associated with an increase in seizures, and this may increase the sleep complaints. Medications and increased psychological stress may contribute to sleep disruption. Also, during the later stages of pregnancy, women may develop hypoventilation from compression of the diaphragm. For women who develop sleepiness, unrefreshing sleep, snoring, or morning headache, a polysomnogram should be considered. CPAP can be effectively used in pregnant women. All women of childbearing potential should be placed on folate, 1 to 4 mg per day, which reduces the risk of fetal malformation.

Anesthesia

Patients with epilepsy undergoing procedures requiring anesthesia should be allowed to continue to take their medication unless there is a clear contraindication. This is to decrease the likelihood of seizures from the sudden withdraw of medication. Anesthesiologists frequently question patients prior to anesthesia for symptoms of sleep apnea, and the popular STOP-BANG questionnaire is a frequently used tool to identify at-risk patients (46). This questionnaire has not been validated in patients with epilepsy.