Since antiquity, physicians have cautioned their patients with epilepsy to avoid sleep deprivation.²¹ Hippocrates (fifth century bc) emphasized that a patient prone to epileptic seizures should “spend the day awake and the night asleep. If this habit be disturbed, it is not so good…worst of all when one sleeps neither night nor day.”²² To avoid recurrence of seizures, Soranus (second century) cautioned his patients “sleep must be undisturbed” because “on slight impulse the body repeats what it just seems to have abandoned.”²³

By late antiquity and the early Middle Ages, little distinction was made between epilepsy, sleepwalking, madness, and demonic possession. Those with such symptoms were lunatics, suffering from “diseases of the moon,” and prone to attacks recurring at periodic intervals, particularly at night (descriptively called mondsüchtig in German).²⁴ The waxing moon heated the atmosphere, melted the brain, and provoked these attacks.²⁴

Physicians of the nineteenth century continued to debate the susceptibility of epileptics to the moon and its cycles. The German physician Romberg²⁵ (1853) argued for the planetary influence of the moon (especially the new and full phases) on epilepsy. Lunar cycles of the moon as the cause for epileptic seizures was largely laid to rest in 1854 after Moreau showed that daily seizure frequencies for more than 5 years among institutionalized epileptic patients were unrelated. However, research and speculation on the full moon as a causative factor for epileptic seizures has waxed again. Recent studies ponder whether the brightness of a full moon disturbs nighttime sleep, shortening sleep duration and quality, thus provoking seizures. Roosli and colleagues²⁶ (2006) found average nocturnal sleep duration was 19 minutes less and subjects felt more tired when the moon was full compared with the new moon in 31 healthy Swiss adult volunteers who kept sleep diaries for 6 weeks. Baxendale and Fisher²⁷ (2008) found a significant correlation between the mean number of seizures and the fullness of the moon in 1571 seizures recorded in a dedicated epilepsy inpatient unit over 341 days. However, the correlation disappeared when they controlled for the local clarity of the night sky, prompting the investigators to suggest that the brightness of the sky rather than the fullness of the moon is the pathologic factor.

Artificial nighttime light pollution and stellar visibility may lessen nocturnal sleep duration and quality.²⁸ Sixty percent of the world population now sleep under light-polluted skies (>90% in the United States and Europe).²⁸ The nighttime sky on a moonless night far from the city lights has an illuminance ranging from 1 to 5 × 10⁻⁴ lux, in contrast to 0.1 to 0.3 lux when the moon is full. Artificial lighting of the infrastructures we build lights our nighttime skies far more than a full moon. Artificial lighting of shopping centers is often 10 to 20 lux (200 times brighter than natural night illuminance). Research has shown as little as 1.5 lux can affect circadian rhythms. Even a bedroom nightlight, particularly of blue light at a wavelength of 460 nm, has been shown to reduce and delay nocturnal pineal melatonin production.

On the subject of environmental stimuli that reduce nocturnal sleep quality and duration, noise exposure should be considered.²⁹ Studies have shown that sleeping in a very noisy environment can lead to increases in time in bed awake, the number of awakenings, sleep stage shifts, and NREM 1 and 2 (at the expense of NREM 3 and rapid eye movement).^29,30 Noise exposure during sleep may increase heart rate, blood pressure, and body movements.³⁰ The World Health Organization guidelines recommend a maximum sound level of 30 dB for continuous background noise, and 45 dB for individual noise events, to promote good sleep. Reducing nighttime noise to 35 to 45 dB (from 60 to 80 dB) and sound masking in intensive care units have been shown to lead to better outcomes and shorter hospital stays.³¹

Patients with epilepsy are much more likely to complain of insufficient or poor-quality nocturnal sleep than sex-matched and gender-matched healthy controls.^32–42 Inadequate sleep may lead patients with epilepsy to a state of chronic partial sleep deprivation. Many, but not all, studies report that sleep maintenance insomnia and excessive daytime sleepiness (EDS) occur more frequently in adults with epilepsy than in the general population.^36–40 A prospective study using a clinical interview and a standardized sleep questionnaire found 30% of 100 adults with epilepsy to report sleep complaints compared with 10% of 90 controls.³⁸ The adults with epilepsy were more likely to report symptoms suggestive of sleep maintenance insomnia (52% vs 38%), sleep-onset insomnia (34% vs 28%), EDS (19% vs 14%), restless legs (18% vs 12%), and sleep apnea (9% vs 3%) than were controls.³⁸ Similarly, sleep complaints were 2-fold higher (39% vs 18%, P<.0001) among 486 adults with partial (localization-related, focal) epilepsy who responded to a mailed questionnaire in comparison with controls.³⁶ Another prospective case series found that 25% of 124 consecutive adults with epilepsy who visited an outpatient epilepsy clinic over a 10-month period complained of insomnia and 17% complained of EDS.³⁹

Whether the IED activation produced by TSD is due to sleep itself (greater amounts of sleep recorded, sampling effects) or because TSD exerts an independent activating effect has been intensely debated for more than half a century. Some have argued that TSD does not offer greater activation than sleep alone, whereas others think TSD activates IEDs independent of sleep induction.

Mattson and colleagues⁴³ (1965) were the first to systematically study this process. After 26 to 28 hours of wakefulness, IEDs were seen in 34% of 89 subjects who had at least 1 seizure and a normal routine EEG, 56% of 34 patients with convulsive epilepsy and IEDs on routine EEGs, and 0% of 20 patients with neurologic disorders other than epilepsy.⁴³ Rowan and colleagues⁴⁴ (1982) found a significantly greater IED yield after TSD compared with routine wake and drug-induced sleep EEGs; IEDs were recorded in 28% of their subjects only after TSD and activated a new epileptic focus in 7%. Degen and Degen,⁴⁵ who spent years studying the effects of sleep deprivation on IEDs, found that (1) for most seizure types, spontaneous sleep and sleep-deprived recordings produced similar activation rates; and (2) seizures were more likely to be activated by sleep or sleep deprivation in patients who had idiopathic primary generalized epilepsy rather than partial focal (localization-related) epilepsy. A prospective study of 721 subjects who had a second EEG after the first was inconclusive found a significantly greater percentage containing IEDs after TSD as compared with a second routine record (23% vs 10%).⁴⁶

Studies in adults suggest that sleep deprivation remains an easy and cost-effective strategy to increase the likelihood of recording IEDs. Leach and colleagues⁴⁷ (2006) systematically evaluated the diagnostic yield of sleep deprivation in 85 patients. Generalized spike-wave discharges were seen in 36 patients (43%) on at least 1 EEG, and focal discharges in 15 (18%) patients. The sensitivity of sleep deprivation was 92%, 58% for drug-induced EEG and 44% for routine EEG. Among the 15 patients showing focal discharges, sleep-deprived EEG provoked abnormalities in 11 (73%) patients. Routine and drug-induced EEG produced abnormalities in 40% and 27%, respectively. Seven (47%) patients had changes seen only after sleep deprivation. Only 2 (13%) patients had IEDs only on the routine EEG, and 1 patient had IEDs only on the drug-induced sleep recording. The authors argued that sleep deprivation was an easy and inexpensive way of increasing the yield of IEDs in young patients presenting with epilepsy.

Gandelman-Marton and Theitler⁴⁸ (2011) found that IEDs were recorded far more often if a sleep-deprived EEG was done within 3 days after a first seizure in 78 adults whose first routine EEG showed no IEDs. Oldani and colleagues⁴⁹ (1998) evaluated the reliability of routine EEG, daytime EEG after sleep deprivation, and nocturnal video-polysomnography (PSG), to diagnose nocturnal frontal lobe epilepsy in 23 patients. All patients had normal video-EEG when awake. Nocturnal video-PSG confirmed the diagnosis in 87% of patients and daytime video-EEG, with sleep deprivation in 52%. Labate and colleagues⁵⁰ (2007) found that generalized IEDs were much more likely to occur on the routine non–sleep-deprived EEG if the study was recorded at 9 am rather than 3 pm in 29 patients with juvenile myoclonic epilepsy (JME).