Sleep and epilepsy: what we know, don't know, and need to know

Influence of sleep and sleep deprivation on ictal and interictal epileptiform activity

Sleep is probably one of the most important physiological factors implicated both in epileptic seizures and interictal epileptiform discharges. The neurophysiology concerning the relationship between sleep and epilepsy is well described in the literature; however, the pathological events that culminate in the seizures are poorly explored. The present paper intends to make a rigorous approach to the main mechanisms involved in this reciprocal relation. Knowledge of sleep and sleep deprivation effects in epilepsy stands as crucial in the understanding of how seizures are produced, their possible lines of treatment, and future research.

Effects of zonisamide as add-on therapy on sleep-wake cycle in focal epilepsy: a polysomnographic study

PURPOSE

The purpose of this study was to evaluate the effects of zonisamide (ZNS) as adjunctive therapy on sleep-wake cycle and daytime somnolence in adult patients affected by focal epilepsy.

METHODS

Thirteen patients affected by focal epilepsy were recruited to undergo a 24-hour ambulatory polysomnography, Multiple Sleep Latency Test (MSLT), and a subjective evaluation of nocturnal sleep by means of the Pittsburgh Sleep Quality Index (PSQI) and daytime somnolence by means of the Epworth Sleepiness Scale (ESS) before and after 3 months of treatment with ZNS as add-on therapy.

RESULTS

Twelve patients completed the study. Zonisamide therapy reduced seizures by >50% in 8 out of 12 patients. Zonisamide did not induce any significant changes in nocturnal polysomnographic variables and in PSQI scores. In addition, mean sleep latency and ESS score were unmodified after treatment.

CONCLUSION

Zonisamide seems to be effective and safe in focal epilepsy. Both subjective and objective sleep parameters showed no detrimental effects on nocturnal sleep and daytime somnolence in patients with focal epilepsy using ZNS. Since some AEDs induce sleep impairment, which is known to trigger EEG abnormalities and seizures and to worsen quality of life, our findings suggest a positive profile of ZNS.

Sleep loss as risk factor for neurologic disorders: a review

Sleep loss refers to sleep of shorter duration than the average baseline need of seven to eight hours per night. Sleep loss and sleep deprivation have severe effects on human health. In this article, we review the main aspects of sleep loss, taking into account its effects on the central nervous system. The neurocognitive and behavioral effects of sleep loss are well known. However, there is an increasing amount of research pointing to sleep deprivation as a risk factor for neurologic diseases, namely stroke, multiple sclerosis, Alzheimer's disease, headache, epilepsy, pain, and somnambulism. Conversely, sleep loss has been reported to be a potential protective factor against Parkinson's disease. The pathophysiology involved in this relationship is multiple, comprising immune, neuroendocrine, autonomic, and vascular mechanisms. It is extremely important to identify the individuals at risk, since recognition and adequate treatment of their sleep problems may reduce the risk of certain neurologic disorders.

Circadian dentate gyrus excitability in a rat model of temporal lobe epilepsy

In human mesial temporal lobe epilepsy (mTLE), seizure occurrence peaks in the late afternoon and early evening. This temporal binding of seizures has been replicated in animal models of mTLE following electrically-induced status epilepticus (SE). We hypothesized that in chronic epilepsy, alterations of circadian excitatory and inhibitory functions of the dentate gyrus (DG), which is believed to regulate the generation of limbic seizures, pathophysiologically contribute to the temporal binding of ictogenesis. We performed electrophysiological single and paired pulse measurements hourly over 24h in the DG of epileptic rats (n=8) 8 weeks after electrically induced SE. Results were compared to individual data obtained before induction of SE and to those of control animals (n=3). Pre and post SE data were analyzed in two distinct phases of the day, i.e. a high-seizure phase between 2p.m. and 10p.m. and a low-seizure phase between 10p.m. and 2p.m. In chronic epileptic animals, latency of evoked potentials was significantly reduced in the high-seizure phase (p=0.027) but not in the low-seizure phase. Compared to baseline values, paired pulse inhibition was significantly increased during the low-seizure phase (interpulse interval (IPI) 25ms, p=0.003; IPI 30ms; p<0.001) but not in the high-seizure phase. Similarly, when compared to controls, inhibition at IPI 20ms was diminished only in the high-seizure phase (p=0.027). Thus, in chronic epileptic animals, DG excitability is increased in the afternoon and early evening possibly contributing to the time of day-dependency of spontaneous seizures in this model system of mTLE. Alterations of circadian DG excitability in epileptic animals may be influenced by changes in hypothalamus-regulated superordinate functions such as excretion of endocrine hormones but further studies are needed.

Epilepsy and sleep

PURPOSE OF REVIEW

There is a complex relationship between epilepsy, sleep and sleep disorders. Recent studies have provided new insights into the links between the disorders that may facilitate differential diagnosis and treatment but may also improve our understanding of underlying pathophysiological mechanisms.

RECENT FINDINGS

Sleep and sleep deprivation have long been recognized to influence interictal epileptiform discharges and seizures. More recent studies have shown that primary sleep disorders such as obstructive sleep apnoea may worsen epilepsy and treatment of these sleep disorders can lead to improved seizure control. Seizures may interfere with night-time sleep structure and cause excessive day-time somnolence (EDS). Antiepileptic drugs may also cause EDS or influence sleep. Despite more frequent use of video-electroencephalographic telemetry and polysomnography, the differential diagnostic challenges between nonrapid eye movement parasomnia and nocturnal frontal lobe epilepsy remain. There is also ongoing debate regarding the possibility of a common pathogenic background for parasomnias and nocturnal seizures that is summarized in the review.

SUMMARY

Accurate identification and diagnosis of sleep disorders as well as epilepsy is clinically important to ensure optimal treatment of both epilepsy and sleep disorders. Further studies of these nocturnal events may advance our understanding of underlying pathological mechanisms and the complex relationship between sleep and epilepsy.

Sleep characteristics of patients with epilepsy with pure sleep-related seizures

OBJECTIVE

Our aim was to investigate the clinical features and sleep characteristics of patients with pure sleep-related seizures.

METHODS

Patients with pure sleep epilepsy were prospectively enrolled and their clinical, EEG, and MRI findings investigated. The Medical Outcomes Study Sleep Scale (MOS-SS) was administered after receiving consent.

RESULTS

Thirty-nine of 1401 consecutive patients (2.7%) with pure sleep-related seizures were included. Of these, 30 (76.9%) had epilepsy of unknown cause and 7 had epilepsy with known structural etiologies. Twenty-seven patients reported less than one seizure per month and 19 had been seizure free for at least 1 year. Thirty-four patients participated in our MOS-SS study. Comparison of sleep problems between those with epilepsy and healthy controls and between the subgroups with frequent and rare seizures did not reveal significant differences.

CONCLUSION

Patients with pure sleep seizures had mostly undetermined etiology usually with a good prognosis, and this rare condition did not seem to affect their sleep quality.

Characterizing sleep disorders of adults with tuberous sclerosis complex: a questionnaire-based study and review

An adult cohort with tuberous sclerosis complex was investigated for the prevalence of sleep disturbances and the relationship with seizure variables, medication, and psychological functioning. Information on 35 adults was gathered using four questionnaires: Epworth Sleepiness Scale (ESS), Sleep and Epilepsy Questionnaire (SEQ), Sleep Diagnosis List (SDL), and Adult Self-Report Scale (ASR). In addition, clinical, genetic and electrophysiological data were collected. Of 35 respondents, 25 had a history of epilepsy. A subjective sleep disorder was found in 31% of the cohort. Insomnia scores showed a significant positive correlation with obstructive sleep apnea syndrome and restless legs syndrome scores. Significant correlations were found between daytime sleepiness and scores on depression, antisocial behavior, and use of mental health medication. A subgroup using antiepileptic medication showed high correlations between daytime sleepiness, attention deficits, and anxiety scores.

Short-Duration Epileptic Discharges Show a Distinct Phase Preference During Ongoing Hippocampal Slow Oscillations

Non-REM (slow-wave) sleep has been shown to facilitate temporal lobe epileptiform events, whereas REM sleep seems more restrictive. This state-dependent modulation may be the result of the enhancement of excitatory synaptic transmission and/or the degree of network synchronization expressed within the hippocampus of the temporal lobe. The slow oscillation (SO), a ∼1 Hz oscillatory pattern expressed during non-REM sleep and urethane anesthesia, has been recently shown to facilitate the generation, maintenance, and propagation of stimulus-evoked epileptiform activity in the hippocampus. To further address the state-dependent modulation of epileptic activity during the SO, we studied the properties of short-duration interictal-like activity generated by focal application of penicillin in the hippocampus of urethane-anesthetized rats. Epileptiform spikes were larger but only slightly more prevalent during the SO as opposed to the theta (REM-like) state. More notably, however, epileptic spikes had a significant tendency to occur just following the peak negativity of ongoing SO cycles. Because of the known phase-dependent changes in 1) synaptic excitability (just following the positive peak of the SO) and 2) network synchronization (during the negative peak of the SO), these results suggest that it is the synchrony and not the changes in synaptic excitability that lead to the facilitation of epileptiform activity during sleep-like slow wave states.

The relationship between sleep and epilepsy: the effect on cognitive functioning in children

AIM

The purpose of this review was to examine the possible pathophysiological links between epilepsy, cognition, sleep macro- and microstructure, and sleep disorders to highlight the contributions and interactions of sleep and epilepsy on cognitive functioning in children with epilepsy.

METHOD

PubMed was used as the medical database source. No language restriction was placed on the literature searches, and citations of relevant studies in the paediatric age range (0-18 y) were checked. Studies including a mixed population but with a high percentage of children were also considered.

RESULTS

The searches identified 223 studies. One reviewer scanned these to eliminate obviously irrelevant studies. Three reviewers scanned the remaining 128 studies and their relevant citations. The review showed that several factors could account for the learning impairment in children with epilepsy: aetiology, electroencephalographic (EEG) discharges, and persistence and circadian distribution of seizures, etc. EEG discharges may affect cognition and sleep, even in the absence of clinical or subclinical seizures. The sleep deprivation and/or sleep disruption affect the neurophysiological and neurochemical mechanisms important for the memory-learning process, but also influence the expression of EEG discharges and seizures. Learning and memory consolidation can take place over extended periods, and sleep has been demonstrated to play a fundamental role in these processes through neuroplastic remodelling of neural networks. Epilepsy and EEG paroxysms may affect sleep structure, interfering with these physiological functions.

INTERPRETATION

Improvement in the long-term cognitive-behavioural prognosis of children with epilepsy requires both good sleep quality and good seizure control. The antiepileptic drug of choice should be the one that interferes least with sleep structure and has the best effect on sleep architecture--thus normalizing sleep instability, especially during non-rapid eye movement sleep.

Searching for autism symptomatology in children with epilepsy--a new approach to an established comorbidity

The Prevalence of Autistic Spectrum Disorder in Children Surveyed in a Tertiary Care Epilepsy Clinic

Clarke DF, Roberts W, Daraksan M, Dupuis A, McCabe J, Wood H, Snead OC 3rd, Weiss SK

Epilepsia 2005;46:1970–1977.

It is well documented that children with autistic spectrum disorder (ASD) have an increased prevalence of seizures; however, studies have not been done to evaluate the prevalence of ASD in children with epilepsy. This comorbidity is important to define as early diagnosis and intervention in some children with ASD has been shown to improve outcome.

Method

Children with epilepsy seen in a tertiary care epilepsy clinic were evaluated using validated autism screening questionnaires (ASQ). In addition, questions about sleep-related disorders, behavior, seizure characteristics, antiepileptic agents, and body mass index (BMI) were requested. An attempt was then made to determine if there was a correlation between the factors identified and ASD.

Results

Of the 107 questionnaires returned, 97 ASQs were properly completed and used in this study. Approximately 32% of children fit the ASQ criteria for having ASD. Most children had not been previously diagnosed. Worst behavior and daytime sleepiness was seen in those at greater risk ( p < 0.01). Seizures also occurred earlier (approximately 2 years) in children at risk of having ASD.

Conclusion

Though confirmatory diagnostic evaluations are needed, this questionnaire-based study suggests that children with epilepsy are at greater risk of having ASD, and illustrates the need for more clinical vigilance. Behavioral difficulties and daytime sleepiness identified in these children could potentially affect their ability to learn. It is of interest that the age of seizure onset identified in those at greater risk corresponds with the approximate age of regression identified in some children with ASD.

Attention deficit hyperactivity disorder in children with epilepsy

Attention deficit hyperactivity disorder (ADHD) is more frequent in children with epilepsy than in general pediatric population. Several factors may contribute to this comorbidity, including the underlying brain pathology, the chronic effects of seizures and of the epileptiform EEG discharges, and the effects of antiepileptic drugs. Symptoms of ADHD are more common in some specific types of epilepsies, such as frontal lobe epilepsy, childhood absence epilepsy and Rolandic epilepsy, and may antedate seizure onset in a significant proportion of cases. In epileptic children with symptoms of ADHD, treatment might become a challenge for child neurologists, who are forced to prescribe drugs combinations, to improve the long-term cognitive and behavioral prognosis. Treatment with psychotropic drugs can be initiated safely in most children with epilepsy and ADHD symptoms.

Slow oscillation state facilitates epileptiform events in the hippocampus

In mesial temporal lobe (MTL) epilepsy, which typically involves the hippocampus (HPC), epileptiform events are enhanced during slow wave sleep (SWS). It remains unclear how and why the electroencephalographic (EEG) states that constitute SWS might predispose the HPC to this type of pathological activity. Recently our laboratory has described a novel state of deactivated hippocampal EEG activity that occurs during both SWS and urethan anesthesia: the slow oscillation (SO). This activity is characterized by a high-amplitude approximately 1-Hz signal, high synchrony within the hippocampus, and a dynamic coordination with neocortical SO. To assess how this activity state might influence epileptiform discharges, we studied the properties of stimulation-evoked and spontaneous epileptiform events elicited in the HPC of urethan-anesthetized rats. We compared those elicited during the SO to those occurring during the theta rhythm. The average duration but not the amplitude of evoked afterdischarges (ADs) was consistently larger during the SO. In addition, spontaneous epileptiform events were more frequent and of higher amplitude during the SO. Last, the bilateral propagation of both ADs and spontaneous events in the hippocampus was enhanced during the SO. These results imply that the threshold for the generation and propagation of epileptiform activity in the hippocampus is lowered during the SO and that this state may be a seed for the initiation, maintenance, and generalization of MTL epilepsy. Further examination of the pathophysiology of sleep-epilepsy interactions in the HPC will be of benefit for an understanding of the mechanisms, prognosis, and therapy for this form of epilepsy.

Effects of valproic acid on sleep in children with epilepsy

PURPOSE

Parents frequently report increased sleep duration in their children during treatment with valproic acid (VPA). We assessed sleep duration and sleep behavior before and after tapering VPA in children treated for more than 6 months.

METHODS

Sleep variables were assessed by questionnaire, diary, and actigraphy (for 7 consecutive days and nights) before and 8-12 weeks after termination of VPA.

RESULTS

Forty-six children (age range 1.7-17.4 years) completed the study. The questionnaire data showed no significant difference in bed and wake time, duration of sleep, and time to fall asleep before and after ending VPA treatment, although some qualitative measures on daytime sleepiness improved after tapering VPA. The actigraphy data revealed that the average sleep amount without VPA was reduced in 33 children (9 of them >30 min) and longer in 13 children (1 of them >30 min). The mean Assumed Sleep Time per Day decreased by 15.2 min or 9.5 min when the physiologic decrease of sleep duration within 0.3 years was considered. Also mean Actual Sleep Time per Day was significantly reduced after VPA termination (-15.2 min; after correction -10.7 min). The reduction was only significant in children older than age 6 years.

DISCUSSION

Termination of VPA after long-term treatment leads to a significant reduction of sleep duration in children older than 6 years of age. The change was small in the majority, but considerable in a subgroup of children.

The circadian rhythm and its interaction with human epilepsy: a review of literature

Knowledge on the interaction between circadian rhythm and human epilepsy is relatively poor, although if it exists, this interaction may be of value for better knowledge of pathophysiology and for timing of diagnostic procedures and therapy. It appears that human seizure occurrence may have 24-h rhythmicity, depending on the origin. These findings are endorsed by animal studies. Rats placed in constant darkness showed spontaneous limbic seizures occurring in an endogenously mediated circadian pattern. More studies are available on the influence of epilepsy on circadian rhythms. Significant differences in chronotypes between patients with different epilepsy syndromes have been found and numerous studies have described influences of epilepsy and seizures on sleep. In contrast, knowledge on (core) body temperature and clock genes in patients is minimal. Reduced heart rate variability and changed hormone levels, which are under the influence of the biological clock, have been observed in people with epilepsy. In short, large gaps in the knowledge about the interaction of circadian rhythm and human epilepsy still remain. Proposals for studies in this borderline area between the biological clock and epilepsy will be discussed.

Sleep spindle alterations in patients with malformations of cortical development

Malformations of cortical development are disorders of altered brain anatomy and architecture that arise from abnormalities in the usual processes of cerebral cortical development. Although they often lead to epilepsy, cognitive delay, and motor impairment, little is known about their effect on sleep. Since malformations may anatomically or functionally disrupt the cerebral circuits that mediate sleep spindles, we hypothesized that these disorders would be associated with abnormal spindle characteristics. We analyzed the density, maximum frequency, laterality and distribution of sleep spindles seen in routine and long-term electroencephalographic recordings performed in ten brain malformation subjects and ten matched controls. There were no significant differences in spindle density or maximum frequency between the two groups, but malformation subjects had a significantly lower proportion of bilateral spindles and a significantly higher proportion of anterior and diffuse spindles compared to controls. In addition, unilateral malformations appeared to be associated with a skewing of unilateral spindles toward the contralateral side. Our findings suggest that brain malformations disrupt the thalamocortical circuits responsible for sleep spindle generation, and support the need for further studies on the relationships between cortical maldevelopment and sleep.

Circadian rhythms, diet, and neuronal excitability

Many aspects of normal biological function are governed by circadian rhythms, including metabolism, ingestive behaviors, and sleep-wake cycles. Certain pathological processes as well are also affected by the circadian rhythms. For example, it is well known that seizure occurrence can be influenced by the sleep-wake cycle. The ketogenic diet (KD) is a high-fat, low-carbohydrate diet with proven efficacy in the treatment of intractable epilepsies. While the mechanisms underlying the clinical effects of the KD remain unclear, there are emerging links between circadian rhythms and KD action. Such factors should be considered when planning and administering the KD to patients.

Background sodium current underlying respiratory rhythm regularity

Rhythm-generating neural circuits underlying diverse behaviors such as locomotion, sleep states, digestion and respiration play critical roles in our lives. Irregularities in these rhythmic behaviors characterize disease states--thus, it is essential that we identify the ionic and/or cellular mechanisms that are necessary for triggering these rhythmic behaviors on a regular basis. Here, we examine which ionic conductances underlie regular or 'stable' respiratory activities, which are proposed to underlie eupnea, or normal quiet breathing. We used a mouse in vitro medullary slice preparation containing the rhythmogenic respiratory neural circuit, called the preBötzinger complex (preBötC), that underlies inspiratory respiratory activity. We varied either [K(+)](o) or [Na(+)](o), or blocked voltage-gated calcium channels, while recording from synaptically isolated respiratory pacemakers, and examined which of these manipulations resulted in their endogenous bursting becoming more irregular. Of these, lowering [Na(+)](o) increased the irregularity of endogenous bursting by synaptically isolated pacemakers. Lowering [Na(+)](o) also decreased the regularity of fictive eupneic activity generated by the ventral respiratory group (VRG) population and hypoglossal motor output. Voltage clamp data indicate that lowering [Na(+)](o), in a range that results in irregular population rhythm generation, decreased persistent sodium currents, but not transient sodium currents underlying action potentials. Our data suggest that background sodium currents play a major role in determining the regularity of the fictive eupneic respiratory rhythm.

Rhythmic Constraints on Hippocampal Processing: State and Phase-Related Fluctuations of Synaptic Excitability During Theta and the Slow Oscillation

Coordinated patterns of state-dependent synchronized oscillatory activity have been suggested to play differential roles in both the encoding and consolidation phases of hippocampal-dependent memories. Previous studies have concentrated on the mutually exclusive patterns of theta and sharp-wave/ripple activity because these were thought to be the only collective oscillatory patterns expressed in the hippocampus. Recently we (and others) have described a novel rhythmic activity expressed during anesthesia and deep sleep, the hippocampal slow oscillation (SO). In an attempt to describe the differential effects of theta and the SO on processing in the hippocampal circuit, we performed evoked potential analysis of two major pathways (the commissural and perforant) in urethan-anesthetized rats across spontaneously expressed theta and SO states. We show that synaptic excitability was significantly enhanced in all pathways during the SO as compared with theta with the exception of the medial perforant path to the dentate gyrus, which showed greater excitability during theta. Furthermore, within each ongoing rhythm, there was a phase-dependent modulation of synaptic excitability. This occurred across all sites and similarly favored the falling phase (positive to negative) of both theta and the SO. Differential effects on the input, processing, and output circuitries of the hippocampus across mutually exclusive coordinated oscillatory patterns expressed during different states may be relevant for the staging of memory processes in the medial temporal lobe.

Ictal scalp EEG recording during sleep and wakefulness: Diagnostic implications for seizure localization and lateralization

To determine the localizing value of electroencephalography (EEG) for seizures during sleep versus seizures during wakefulness, we compared scalp EEG for 58 seizures that occurred during sleep with 76 seizures during wake in 28 consecutive patients with temporal lobe epilepsy. Regression analysis showed that seizures during sleep are 2.5 times more likely to have focal EEG onset (p = 0.01) and 4 times more likely to correctly localize seizure onset (p = 0.04) than seizures during wake. EEG seizure onset preceded clinical onset by a longer duration in sleep seizures (mean, 4.69 s) than in wake seizures (mean, 1.23 s; p < 0.01). Sleep seizures showed fewer artifacts, but the difference was not significant (p = 0.07). For temporal lobectomy candidates undergoing video-EEG monitoring, the recording of seizures during sleep may be favored.