Sleep, night and SUDEP
[Part 2: Questions; Buchanan G, Li R, Petrucci A & Purnell B]
It has been appreciated for quite some time that SUDEP occurs frequently during the night (Lamberts et al., 2012; Ali et al., 2017). While this is a recurrent scenario in the case vignettes presented in this resource, it is perhaps best exemplified by the Mortality in Epilepsy Monitoring Units study (MORTEMUS) in which 16 SUDEP cases that occurred during epilepsy monitoring were evaluated. Fourteen of these cases occurred during the night. Seven of the 10 cases for which sleep/wake state could be determined occurred during sleep (Ryvlin et al., 2013). There are a variety of factors that may contribute to the nighttime predilection of SUDEP, including circumstances associated with being alone, circumstances associated with being in bed, sleep related factors, and time-of-day or circadian related factors.
Circumstances associated with being alone. When SUDEP happens at night, the person is often found deceased after having been alone and unobserved through the night. It has been suggested that reduced supervision and monitoring during the night, and thus, delayed or absent resuscitative efforts, contribute to the day-night distribution of SUDEP. Early reports indicate that SUDEP occurs less frequently when patients with epilepsy live together. Thus if someone had a seizure, another person was there to attend to them and administer a potentially lifesaving intervention thereby reducing SUDEP risk (Nashef et al., 1995). This hypothesis is corroborated by the findings that more rapid intervention reduces the seizure duration and results in a less drastic reduction of oxygen concentration in the blood (Seyal et al., 2013). In the MORTEMUS study, when resuscitation was attempted, SUDEP was averted (Ryvlin et al., 2013). Increasing nocturnal supervision by the use of monitoring devices, regular checks or having someone else sleep in the same room as the patient is currently the best way of reducing nocturnal SUDEP risk (Ryvlin et al., 2006; Langan et al., 2005; Geertsema et al., 2018; van der Lende et al., 2016).
Circumstances associated with being in bed. Not that surprisingly since people spend the night in bed, patients who die of SUDEP are frequently found in bed, face down. A recent meta-analysis reveals that upwards of 80% of SUDEP victims are found face down, often with their nose and mouth in a pillow or blanket (Ali et al., 2017). This could indicate that the airway was blocked following a seizure leading to a failure to exchange the air in the lungs and a fatal inability to oppose falling levels of oxygen and rising levels of carbon dioxide (CO2) in the blood. The rise in CO2 is initially good, as it increases respiratory drive under normal circumstances (Richerson, 2004), and can cause arousal from sleep (Berthon-Jones & Sullivan, 1984; Buchanan & Richerson, 2010; Smith et al., 2018). However, if the airway continues to be obstructed the abnormalities in blood gases will continue until they reach potentially fatal levels. Similar mechanisms have been proposed for the sudden infant death syndrome (SIDS), another sudden death entity that has similarities to SUDEP (Kinney et al., 2009). In SIDS, the “Back to Sleep” campaign wherein babies are placed on their backs to sleep, resulted in a significant reduction in incidence (Hunt and Brouillette, 1987). A similar concept has been proposed for SUDEP (Tao et al., 2015), but it is unclear if this would be helpful given that the body position will likely change if a seizure occurs. Interestingly, one study found that non-fatal seizures are associated with the patient not lying face down (Shmuely et al., 2016).
Seizures during sleep. Seizures can have profound effects on cardiovascular and respiratory function. At present, SUDEP is thought to occur due to seizure induced disruption of breathing and/or cardiac function (Devinsky et al., 2016). Sleep also affects cardiovascular and respiratory function (Snyder et al., 1964; Shea, 1996). Thus, it follows that the combined effects of sleep and seizures on cardiovascular and respiratory function could be particularly detrimental. Indeed, in at least one animal model, seizures induced during sleep have more profound effects on breathing and on suppression of brain electrical activity. Seizures induced during sleep, especially rapid eye movement (REM) sleep, in this model are more likely to be fatal (Hajek and Buchanan, 2016). Even non-fatal seizures induced during non-REM (NREM) sleep suppress breathing to a greater degree than seizures induced during wakefulness. Seizures induced during NREM sleep are also associated with prolonged suppression of EEG activity following the seizure in this model (Hajek & Buchanan, 2016) and in patients in an epilepsy monitoring unit (Peng et al., 2017). EEG suppression may correlate with reduced arousability following the seizure in mice (Petrucci et al., 2018) and duration of respiratory dysfunction in humans (Kuo et al., 2016). The duration of such post-ictal generalized EEG suppression (PGES) has been correlated with an increased risk of SUDEP in at least one study (Lhatoo et al., 2010), but not others (Surges et al., 2011; Kang et al., 2017). Notably, many neurotransmitter systems that have been implicated in SUDEP are modulated in a sleep state-dependent manner and are involved in sleep-wake regulation (Brown et al., 2012; Mitchell and Weinshenker, 2010; Basheer et al., 2004). These include serotonin (Buchanan et al., 2014; Richerson & Buchanan, 2011), norepinephrine (Zhang et al., 2017; Zhao et al., 2017), and adenosine (Shen et al., 2010; Lazarus et al., 2017).
Although seizures induced during REM sleep in the aforementioned animal model are fatal (Hajek & Buchanan, 2016), seizures are generally more likely to occur during NREM sleep than during REM sleep in patients with epilepsy (Ng & Pavlova, 2013) and in many, but not all, animal models of epilepsy (Shouse et al., 2004; Sedigh-Sarvestani et al., 2014b), suggesting an underlying seizure gating mechanism to prevent seizures, and consequently SUDEP. One candidate mechanism for this may involve serotonin. Serotonin modulates both sleep (Monti, 2011) and seizures (Bagdy et al., 2007), and serotonin neurons are essential in CO2-induced arousal (Smith et al., 2018; Buchanan & Richerson, 2010), which may play a critical role in arousal after a seizure. These serotonin neurons project to the hippocampus (Commons, 2016), a key node in the pathogenesis of epilepsy (Huberfeld et al., 2015), and participate in regulation of electrocerebral rhythms during REM sleep (Pignatelli et al., 2012) including those that have been found to have anticonvulsant effect in animal seizure models (Miller et al., 1994).
Seizures during the night. While the fact that SUDEP happens during the night, could speak to effects of sleep, or to being unattended, this could also suggest independent time-of-day, or circadian, influences. Seizures are subject to both sleep-state dependent (Ng & Pavlova, 2013; Sedigh-Sarvestani et al., 2014a) and circadian dependent regulation (Durazzo et al., 2008; Quigg et al., 1998; Loddenkemper et al., 2011). Cardiovascular (Piccione et al., 2005) and respiratory (Buchanan, 2013) function are also regulated in a circadian fashion. Thus, cardiorespiratory changes which occur during the night could increase SUDEP risk independent of sleep state. In the same aforementioned animal model, time-of-day altered the outcomes of seizures both in terms of breathing and EEG suppression (Purnell et al., 2017). In another animal model frequently employed in the study of SUDEP, mortality is heavily influenced by time of day (Moore et al., 2014). Many of the neurotransmitter systems relevant to SUDEP are regulated in a circadian fashion (Poncet et al., 1993; Linsell et al., 1985; Chagoya de Sanchez, 1995). Furthermore, physiological parameters such as Q-T interval and the respiratory response to CO2, which have been implicated in SUDEP etiology, are different during the night (Molnar et al., 1996; Buchanan, 2013; Spengler et al., 2000; Peever & Stephenson, 1997; Bulow, 1963).
We need to better understand how all these factors interact to result in SUDEP. Animal models have been and will continue to be helpful in parsing out the mechanisms of SUDEP. The key to reliably predicting and preventing the occurrence of SUDEP lies in our ability to understand and model SUDEP mechanisms.
Gordon F. Buchanan [1,2], Rui Li , Alexandra N. Petrucci [1,2], Benton S. Purnell [1,2]
 Department of Neurology &  Neuroscience Program
University of Iowa, Carver College of Medicine, Iowa City, IA 52242, USA
How to cite:
Buchanan G, Li R, Petrucci A & Purnell B. Sleep, night and SUDEP. In: Hanna J, Panelli R, Jeffs T, editors. Continuing the global conversation [online]. SUDEP Action & SUDEP Aware; 2018 [retrieved day/month/year]. Available from: www.sudepglobalconversation.com.