Clinical manifestations

Presentation and course

Certain types of headaches may be associated with specific stages of sleep and thereby lead to sleep disruption. Migraine attacks have been described in association with sleep stages 3, 4, and REM. Cluster headaches and chronic paroxysmal hemicrania have been linked with stage REM. Sleep-related migraine attacks similar to daytime episodes are characterized by unilateral throbbing head pain associated with nausea and vomiting, scotomata, visual field defects, photophobia, paresthesias, and sometimes hemiparesis or aphasia. Migraines are highly idiosyncratic and not all symptoms are present. Attacks last several hours to a few days.

Cluster headaches that commonly occur during sleep are characterized by severe unilateral, periorbital, malar, and temporal pain with lacrimation, nasal engorgement, rhinorrhea, forehead sweating, and flushing of the malar area. Attacks last no more than 2 hours and may occur several times daily, sometimes at the same time of the day or night, commencing and terminating abruptly. Spontaneous remissions lasting several months characterize the syndrome. Transient recurrent situational insomnia in association with cluster headache resolves after the headache syndrome subsides (90). Chronic insomnia and shift work seem to be common among Arctic cluster headache patients. In a small study of Arctic workers, insomnia was significantly associated with shift work and longer-lasting cluster bouts (70). Eighty percent often had headache attacks during sleep, particularly between 24:00 am and 04:00 am. Shift workers were significantly more likely to see lack of sleep as a cluster attack trigger than daytime workers, supporting the idea of cluster headache as a circadian rhythm disorder.

In chronic paroxysmal hemicrania, attacks of severe pain are shorter than with cluster headache, but they are more frequent and are also associated with conjunctival hyperemia, rhinorrhea, and more rarely Horner syndrome. They may occur predominantly at night, usually at the same hour, and do not remit spontaneously. The sometimes close linkage to REM sleep has led to the descriptive term "REM sleep-locked headache."

Headaches may occur on awakening with a variety of disorders. Although common in children with brain tumors, they appear in only 5% of adults with brain tumors. In the sleep apnea syndrome, patients often complain of diffuse headache in the morning localized to the frontal region, but the incidence is not related to the severity of the disease (02). The intensity of pain is mild to moderate and tends to disappear shortly after getting out of bed. Morning headache may not be more common in patients with sleep apnea syndrome than in other sleep disorders, but the successful treatment of sleep apnea is associated with significant improvement of chronic headache in 30% of patients (79). In patients with sleep apnea syndrome, prolonged afternoon naps may also be followed by headache. Headaches on awakening in the morning may also occur in relation to bruxism, systemic hypertension, depression, muscle-contraction headache, alcohol intoxication, and sinus inflammation. In a study of 721 consecutive patients comparing the characteristics of idiopathic intracranial hypertension in men and women (09), the authors found that men were more likely to have sleep apnea (24% vs. 4%, p < 0.001) and were older (37 vs. 28 years, p = 0.02). Men were less likely to report headache (55% vs. 75%, p < 0.001) as their first symptom of intracranial hypertension but had more visual disturbances (35% vs. 20%, p = 0.005).

Chronic headaches in children may be seen in association with decreased duration of sleep at night, poor sleep hygiene, fragmentation of sleep, increased number of somnambulism, somniloquy, and enuresis events, and nocturnal snoring. A study found that children with headaches have more excessive daytime sleepiness, narcolepsy, and insomnia than children without headaches (p< 0.005), whereas sleep apnea and parasomnias may be more specific for genuine migraines and not for headaches in general (61). Pediatricians should inquire about daytime sleepiness, narcolepsy, and insomnia in children with headaches. In a polysomnographic study of children with headaches, the authors found results supporting an association between migraine and sleep-disordered breathing, and between tension headache and bruxism (100). Reduced REM and slow-wave sleep in patients with severe and chronic migraine headache suggests an intrinsic relationship between sleep and headache disorders. In a study, the relative risk of exhibiting primary headache was 13.1 among children with sleep bruxism whose parents were separated (08). In another study, a retrospective chart review of 527 patients, ages 7 to 17 years, with a primary headache diagnosis of migraine (n = 278), tension-type headache (TTH: n = 157), and new daily persistent headache (NDPH: n = 92), sleep disturbance was greater in patients with tension-type headache (10.34 ± 5.94, P = .002) and new daily persistent headache (11.52 ± 6.40, P <  .001) than migraine (8.31 ± 5.89) (83).

Hypnic headache is a benign headache disorder of the late middle-aged and the elderly, rarely of children, characterized by regular awakenings from sleep at a constant time of night (03; 59). The headache is generally diffuse and bilateral, lasting 30 to 60 minutes, and is sometimes associated with a dream. Dodick and colleagues reported 19 cases with a mean age of headache onset of 60.5 + 9 years (26). Patients were generally awakened between 1 AM and 3 AM; three patients reported that a similar headache could occur during daytime naps as well. In 68% of patients with headache, frequency was 4 nights per week, with resolution occurring within 2 hours. As more patients are described, the number of unilateral hypnic headaches reported has increased. In a review of 71 cases, Evers and Goadsby described the following characteristics: age at onset, 63 years, female 63%; duration of attack, 67+ 44 minutes; frequency of attacks, 1.2 per 24 hours; moderate intensity of pain 67%, bilateral pain 61%, diffuse pain 56% (30). Hypnic headache is rare in childhood; its clinical features are inconsistent, but in a review study, autonomic symptoms were absent in all patients. Successful treatment with melatonin was reported in two of five patients (95). In a retrospective study of 40 patients (80% females) with hypnic headache and mean follow-up of 929 days, the average age of headache onset was 62 years (range 44 to 86) (97). Twenty (50%) patients had previous history of migraine, 5% had bilateral cranial autonomic features, and 40% underlying sleep abnormalities. The average duration per day and frequency per month of headaches were 186 minutes (range 30 to 720 minutes) and 21 days (range 5 to 30), respectively. The best response was obtained with lithium (7 of 10 complete response and 2 of 10 moderate response). Caffeine induced a complete response in 6 of 21 and a moderate response in 9 of 21 subjects. In another review of cases published from 1988 to 2018, the authors reviewed 343 adults (69.0% women, 31.0% men) and five children diagnosed with hypnic headache (96). The average age of adults was 58.0 ± 13.1 years, ranging from 15 to 85 years, and the average age of children was 9 years, ranging from 7 to 11 years. The diagnosis was made 7.6 ± 14.2 years after onset of headache. Pain occurred during nocturnal sleep in 94.8%, with an average duration of 90 minutes, and was bilateral in 55.5%. Headache was dull in 74.4% and of moderate intensity in 61.5%; 94.5% of the patients reported headache occurring 10 or more days per month. Autonomic manifestations occurred in 7.6% of the patients, predominantly lacrimation and rhinorrhea. Caffeine provoked the best therapeutic response in acute treatment. Prophylaxis was successful with lithium, caffeine, and indomethacin in 77.8% of the patients. Treatment produced remission in 56.7% of the patients, and 72.7% had no recurrence.

The exploding head syndrome or “snapping of the brain” (episodic cranial sensory shock) is characterized by flashing lights and sounds during the night that terrify patients (40). Pain is not present. Polysomnographic studies have shown that the attacks take place during all stages of sleep including REM sleep, without evidence of epileptogenic discharges. There is one case report of exploding head syndrome followed by sleep paralysis and migraine (27).

Hemicrania horologica or clock-like hemicrania is a rare disorder with headaches that occur with clock-like precision every 60 minutes, day and night, lasting 15 minutes (39). Hemicrania horologica differs from chronic paroxysmal hemicrania in the lack of autonomic signs, the clock-like regularity over 24 hours, and the response to nonsteroidal anti-inflammatory drugs in addition to indomethacin. Unlike hypnic headache, attacks also occur during the day.

Clenching and grinding of teeth during sleep, or bruxism, is a sleep-related movement disorder occurring predominantly in stages 1 and 2 of sleep, although it has also been noted in REM sleep. Hundreds of events occurring during the night may lead to abnormal wear of the teeth, temporomandibular joint disorder, and jaw pain. Some patients with bruxism have associated muscle-contraction headaches. In a study of carefully classified patients, the authors identified a clear relation between muscle-contraction headaches and temporomandibular symptoms, depression, anxiety, poor sleep, and impairments of oral function (13). The authors concluded that the findings indicate a close, but incomplete, overlap between muscle-contraction headaches and temporomandibular joint disorder, advocating for further research. In a study, children with sleep bruxism demonstrated a greater risk of having primary headache, and those whose parents were separated had a greater chance of having headache. Sleep bruxism was associated with headache, but clenching the teeth during waking hours was not correlated with primary headache (08).

Chronic insomnia and headaches along with memory deficit, depression, anxiety, irritability, and judgment and personality disorder form part of the post-concussion syndrome that develops following head trauma in susceptible patients. Polysomnographic studies immediately following severe head trauma have shown alpha-wave intrusion, increased amount of sleep spindles, decreased REM density, and enhanced muscle tone in REM sleep, while individual sleep stages become less distinct. Weeks to months following recovery of consciousness, stages 3, 4, and REM are decreased along with total sleep time. Sleep spindles are decreased and the polysomnogram is punctuated with numerous arousals and awakenings. The post-concussion syndrome appears within hours to days following head trauma and may persist for many months thereafter.

Self-reported snoring was associated with headache (odds ratio 1.5 with a 95% confidence interval) independent of other potential confounders in a study of 3323 middle-aged and elderly men in Denmark (47). Although obstructive sleep apnea could underlie such an association, lack of nocturnal measurements precluded the evaluation of that hypothesis.

Insomnia and daytime fatigue are common complaints in patients with chronic headache. Fibromyalgia occurs in 35.6% of patients with chronic migraine, also known as transformed migraine, and this group of patients has a higher incidence of insomnia (77). In a study of 784 subjects, the prevalence of headache and headache frequency were higher in subjects referred for polysomnography for any sleep disturbance independently of obstructive sleep apnea, compared to general population controls (06). History of repeated concussion is associated with higher reported sleep disturbance, more severe headaches, mood disturbance, and cognitive dysfunction in children, adolescents (103), and adults (72).

Drugs used for treatment of sleep disorders may have headache as a common adverse event. Safety and tolerability data of modafinil from six double-blind, placebo-controlled studies evaluated 1529 outpatients randomized to modafinil 200 mg, 300 mg, 400 mg, or placebo once daily for up to 12 weeks. Headache appeared in 34% of patients receiving modafinil versus 23% receiving placebo (86). Headache is also a common adverse effect in patients taking ramelteon for treatment of insomnia and zolpidem extended-release 12.5 mg. Headache is the most common side effect reported by patients taking melatonin also for insomnia. Long-term treatment with amphetamines of patients with narcolepsy may provoke headache along with irritability, bad temper, and profuse sweating.

Chiari type I malformation is characterized by herniation of the cerebellar tonsils through the foramen magnum (31). This is associated with brain stem, medullary, and cranial nerve compression phenomena. The most frequent symptoms are occipital headaches and dizziness. In addition, patients may have less well-known symptoms, including sleep-related respiratory abnormalities. Some studies report a 50% prevalence of sleep apnea-hypopnea syndrome (SAHS), probably associated with sudden death, in some cases of Chiari type I malformation. The identification of sleep apnea may lead to better treatment and improved quality of life.

Concerns regarding the side effects of the new COVID-19 mRNA-1273 vaccine have been raised. In a randomized, cross-sectional study using an independent online survey questionnaire, 87.8% (1116 of 1271) provided complete responses (49). Of them, 38.7% (432 of 1116) received the mRNA-1273 vaccine, and 89.35% were females. Four hundred twenty five of the 432 mRNA-1273 vaccine recipients (98.34%) reported at least one or more symptoms. One hundred and eight of 432 (25%) patients were temporarily unable to perform daily activities, Among all the symptoms reported headache and decreased sleep quality were included. Other symptoms were myalgia, chills, fever, nausea, joint pains, sweating, localized swelling at the injection site, dizziness, itching, rash, decreased appetite, muscle spasm, and brain fogging/confusion. Most of the symptoms reported were non-life threatening. In one study of 87 patients with confirmed SARS-CoV-2, 63 (72.4%) were female, and 65 (74.7%) were white. The mean age was 49.2 years (SD=14.9) (41). The most prevalent symptoms after COVID-19 infection were fatigue, "brain fog," headache, anxiety, and sleep issues.

Prognosis and complications

Sleep-related migraine, cluster headache, and chronic paroxysmal hemicrania are benign conditions that do not alter the lifespan of the individual. Work incapacity may occur during severe attacks, and global incapacity may occur if attacks are severe and frequent. In women, attacks may cease after menopause. Bruxism may lead to excessive and abnormal wear of the teeth, temporomandibular joint degeneration, and chronic headaches.

In one study of patients with episodic migraine, poor sleep was associated with a higher rate of headache recurrence over the ensuing 6 weeks, particularly if there was coexisting moderate or high stress (101).

In an observational, cross-sectional study of 50 patients with chronic migraine, the authors found that predictive factors for sleep quality were depressive symptoms, headache-related disability, and pain catastrophizing (34). Patients with poorer sleep quality had higher levels of pain catastrophizing and depressive symptoms.

Clinical vignette

A 55-year-old man was evaluated in the clinic because of headaches that were occurring at night and on awakening. Several nights of the week he would wake up between midnight and 2 AM with a bifrontal headache that would disappear on getting out of bed. Most mornings he had a diffuse headache that tended to dissipate after breakfast. He also complained of loud snoring and excessive daytime sleepiness that he attributed to the awakenings caused by headaches. The patient was a physician. His colleagues had noted recent episodes of confusion that they felt were the consequence of sleepiness. On examination, he had an acromegaloid facies. Deep tendon reflexes were increased, and plantar responses were equivocally extensor on both sides. A polysomnographic study followed by a multiple sleep latency test showed a respiratory disturbance index of 28 events per hour and mild episodes of desaturation predominantly during REM sleep. Nocturnal REM sleep latency was 25 minutes. The multiple sleep latency test showed presence of REM sleep in the first nap and an average sleep latency of 3 minutes in four nap segments. A diagnosis of sleep apnea disorder and possible narcolepsy was made.

A CT scan of the head ordered by the neurologist and sleep specialist to further evaluate the abnormalities observed in the neurologic examination showed a mass compressing the floor of the third ventricle that the radiologist interpreted as a craniopharyngioma. The patient was referred for neurosurgical management that confirmed the radiological diagnosis.

Causes for concern in this patient are nocturnal headaches that improved on getting out of bed, episodes of confusion, and associated signs of neurologic involvement manifested by increased deep tendon reflexes and equivocal responses to plantar stimulation. Tumors of the floor of the third ventricle may cause excessive daytime sleepiness with REM sleep abnormalities in the polysomnographic study suggestive of narcolepsy. Mild acromegaly as a result of hypothalamic-pituitary dysfunction caused enlargement of the tongue and soft palate with sleep apnea disorder of moderate intensity that was a contributing factor to daytime sleepiness.

Biological basis

Research work within the last 10 years points at the hypothalamus as an early mediator in the pathophysiology of migraine and has revealed mediating signaling molecules such as serotonin and dopamine that are common with sleep.

The hypothalamus is involved in homeostatic regulation, including pain processing and sleep regulation, and is a key integrator of circadian entrainment to light. This is in part regulated by pituitary adenylate cyclase-activating peptide, a peptide that has been extensively studied in relation to headache in both humans and animals. Given its prominent role in circadian entrainment further studies investigating its effect on sleep disturbances and comorbidities in humans are warranted (45).

Furthermore, the discovery of the glymphatic system, a CNS waste removal system, points to a shared pathophysiology manifesting in migraine and sleep problems (102). Some studies have suggested that glymphatic dysfunction might be a common underlying etiology of sleep disorders and headache pain. A review of the current literature on the relationship between the glymphatic system, sleep, and headaches discusses their roles (105).

Etiology and pathogenesis

The pathophysiology and pharmacology of headache and sleep disorders involves a group of neural networks that likely underlie a shared clinical association. The obvious brainstem, hypothalamic, and thalamic convergence suggests a bidirectional influence, but it is difficult to differentiate between cause and effect, whereas a spurious relationship cannot be ruled out (44). The cause of migraine, cluster headaches, and chronic paroxysmal hemicrania remains unknown, and the interaction between headaches and sleep is poorly understood. In rare instances, intracranial tumors cause severe headaches that awaken the patient at night. Meningitis causes insomnia because of head pain.

Migraine headaches may be provoked by sleep, but the most common association is sleep following a migraine attack. The "therapeutic" need for sleep in some patients having an attack of migraine may be related to the disorders of serotonin metabolism in migraine, but proof is lacking. Neural stimuli originating in the hypothalamus or brainstem may provoke changes in cerebral and extracranial circulation. The trigeminovascular system that promotes vasodilatation and release of calcitonin gene-related peptide and substance P has been implicated in the mechanism of migraines because calcitonin gene-related peptide is elevated in the jugular venous blood of migraineurs during the attack (36). Serotonin (5-HT) is released from platelets during migraine headaches, and 5-hydroxyindoleacetic acid, the main metabolite of serotonin, is excreted in excess in the urine following a migraine attack. Sumatriptan (an agonist of the 5-HT1 receptor found in cerebral arteries, where it has an inhibitory effect) aborts the migraine headache. Methysergide (antagonist of the 5-HT2 receptor found mainly in temporal arteries, where it has an excitatory effect) also terminates migraines. Serotonin, implicated in mechanisms of NREM sleep, is a possible neurotransmitter link between migraine and sleep. In a study of 1283 migraineurs drawn from 1480 consecutive headache sufferers presenting for evaluation to a tertiary headache clinic, Kelman and Rains found that migraineurs were 84% female, with a mean age of 37.4 years (52). Groups were formed based on patient's average nocturnal sleep patterns, including short, normal, and long sleep groups, and were compared on headache variables. Sleep complaints were common and associated with headache in a sizable proportion of patients. Over half of migraineurs reported difficulty initiating and maintaining sleep at least occasionally. The short sleep group, who routinely slept 6 hours per night, exhibited the more severe headache patterns and more sleep-related headache. Sleep complaints occurred with greater frequency among chronic than episodic migraineurs. The authors indicated that prospective studies are needed to determine if normalizing sleep times in the short sleepers would impact headache threshold.

In a study, migraineurs showed a lower cyclic alternating pattern rate in NREM sleep and, in particular, a lower number of A1 phases (low-frequency, high-amplitude EEG bursts) compared with the controls (21). Migraineurs also showed a lower index of high-frequency EEG arousals during REM sleep. The reduction in cyclic alternating pattern rate was interpreted by the authors as a lower level of arousal fluctuation in NREM sleep. They suggested that the reduced arousal index in REM was due to a dysfunction in neural structures involved in both the control of REM sleep and the pathophysiology of migraine, including the hypothalamus and the brainstem. Another sleep structural change was found in a retrospective study of 240 patients who had polysomnography recorded in the sleep unit or at home (85). Mean objective sleep latency (29.5 minutes, SD 35.5) was significantly lower than subjective sleep latency (37.4 minutes, SD 41.6) (p< .001). There were significantly more patients with migraine in the overestimation group than in the underestimation group (20 of 159 patients vs. 3 of 81 patients; p=0.035), a difference that was not observed in patients with headache without migraine.

Conclusions. We hypothesize that migraine is markedly associated with overestimation of sleep latency. This overestimation was not observed in patients with other headache types.

The trigeminal autonomic cephalalgias are a group of primary headache disorders that includes cluster headache, paroxysmal hemicrania, and short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing and cranial autonomic features (SUNCT/SUNA) (35). Hemicrania continua is often included with this group. Cluster headache shows a remarkable periodicity in its occurrence, suggesting a linkage to the circadian rhythm. The neurovascular hypothesis suggests excitation of autonomic fibers of the greater superficial petrosal nerve that would be responsible not only for lacrimation and conjunctival injection, but also for edema of the wall of the internal carotid artery with pain and ipsilateral Horner syndrome. The association between migraine attacks, cluster headache, chronic paroxysmal hemicrania, and REM sleep remains unexplained. Autonomic dysfunction was suggested in a 9-week actigraphic recording with repeated polysomnography of a patient with cluster headache, evaluating both sleep macrostructure and microstructure (20). During the acute bout the authors observed an irregular sleep-wake pattern and abnormalities of REM sleep. After the cluster phase these alterations remitted. The authors concluded that cluster headache was associated, at least in this patient, with sleep dysregulation involving the biological clock and the arousal mechanisms, particularly in REM sleep. They hypothesized that the abnormalities were consistent with posterior hypothalamic dysfunction.

Patients with cluster headache have a higher prevalence of sleep apnea, and a possible relationship between these two conditions has been proposed (38). However, there is evidence that cluster headache and obstructive sleep apnea are not causal, but rather parallel processes possibly generated in the hypothalamus. The exact role sleep apnea plays in the perpetuation or precipitation of cluster headache is still to be determined. A study showed that 50% of cluster headache patients were blood pressure nocturnal “non-dippers,” a frequency higher than expected (92). Non-dipper patients displayed higher mean nighttime systolic and diastolic blood pressure. The high incidence of non-dipper pattern in cluster headache and obstructive sleep apnea supports the hypothesis of a relationship between these two disorders. Furthermore, there is a difference between sexes. In a study of 351 patients with cluster headache (male:female ratio of 2:1), the diurnal variation of attacks showed peak prominence in men's attack cycle to be advanced by 1 hour compared to women's, despite no difference in self-reported bedtime. The onset of cluster headache decreased with increasing age for both sexes, but chronic cluster headache was more prevalent in women compared to men (44.0% vs. 31.9%, p < 0.05). The authors conclude that the findings suggest differences in influence of the hypothalamus on attack occurrence between the sexes (62).

In a study, episodic cluster headaches were studied using actigraphy continuously for 2 weeks along with sleep diaries (63). The authors found that during bouts patients (n = 17, 2.3 attacks/day) spent more time in bed (8.4 vs. 7.7 hours, p = 0.021) and slept more (7.2 vs. 6.6 hours, p = 0.036) than controls (n = 15). In remission (n = 11) there were no differences compared with controls. In five patients, attacks occurred at the same hour several nights in a row. Thus, actigraphy offers the possibility of a continuous and long study period in a natural environment.

Sleep quality was reduced in patients with cluster headache compared to controls. "Poor sleepers" were 71.4% (105/147) in chronic cluster and 48.3% (235/367) in patients with episodic cluster headache compared to 25.6% (94/367) in controls (19).

Chronic paroxysmal hemicrania is considered a variant of cluster headache, however, with sufficient individual traits to be listed separately. In this condition, attacks of pain are more frequent, but of shorter duration, than in typical cluster headache, and the therapeutic response to indomethacin is quasi-specific. There is a case report of a patient with headaches meeting the criteria of chronic paroxysmal hemicrania who was fully responsive to indomethacin administration during the first 3 months of treatment (93). Further investigations revealed a macroprolactinoma, and headaches stopped after cabergoline treatment. The authors suggested that patients with paroxysmal hemicrania should be investigated for pituitary abnormalities.

The following mechanisms have been implicated in headaches on awakening in patients with sleep apnea syndrome: hypoxemia, hypercapnia, altered cerebral blood flow, excessive neck movements, and sleep disturbance caused by depression. However, the exact mechanism of headache and the relationship between obstructive sleep apnea, headache, and morning headaches in particular, remain controversial (80). In a study of 462 patients with obstructive sleep apnea compared to controls, the authors found that the apnea-hypopnea index (AHI) was significantly higher in obstructive apnea patients with morning headache compared to patients without morning headaches (37). Morning headache was more frequently reported by women with primary headache history. In the same study morning headache was totally resolved in 90% of patients treated with nasal continuous positive airway pressure.

Based on the regularity of events, the hypnic headache syndrome has been linked to an abnormality of the circadian clock, but scientific proof is lacking. The association with dream content suggests linkage to REM sleep. However, Liang and colleagues showed that hypnic headache occurred equally in both REM and non-REM sleep, and most patients ran an episodic course (58). The clinical spectrum of hypnic headache has been expanded to include unilateral forms (about 40%, half of which are side-locked), forms with a longer duration (up to 3 hours), and cases with onset in juvenile/adult age (22). The location of pain is frontotemporal in over 40% of cases; headache is throbbing in 38% of cases, dull in 57%, and stabbing in less than 5%. Nausea is reported in 19% of cases; photophobia, phonophobia, or both are present in 6.8%. According to some studies, a history of migraine is common in hypnic headache patients, suggesting that hypnic headache and migraine might share a common pathophysiological predisposition (87). One study was unable to confirm a significant role of melatonin concentration changes in the pathophysiology of hypnic headache, discarding melatonin deficiency as a major factor in the pathophysiology of the disorder (66).

Bruxism appears predominantly during stage 2, although occasionally it may be observed in stage REM; its mechanism remains unknown. Headaches associated with bruxism are presumed to be secondary to the muscle activity or the temporomandibular joint stress. There is no satisfactory explanation for the association between childhood migraines and somnambulism.

Of some interest is the report of melatonin-responsive headaches in patients with delayed sleep phase syndrome (67). Headaches of an assorted variety were alleviated with melatonin 5 mg at bedtime in 5 of 30 patients who reported headache with delayed sleep phase syndrome. The authors suggest that its effectiveness may be due to the chronobiological action of melatonin.

An abnormal pattern of hypothalamic hormonal secretion was found in patients with chronic migraine (76). The authors found in 17 patients with chronic migraine and insomnia a decreased nocturnal prolactin peak, increased cortisol concentrations, a delayed nocturnal melatonin peak, and lower melatonin concentrations, which they interpreted as hypothalamic involvement and chronobiological dysregulation due to insomnia. One study provided direct evidence that simultaneous antagonism on both orexin receptors is able to attenuate trigeminal nociceptive activity, as well as induce an elevation of the threshold for the induction of a cortical spreading depression. These data imply that targeting the hypothalamic orexinergic system may offer an entirely novel mechanism for the preventive treatment of migraine with and without aura (43). These findings are supported by the results of another study that found significantly lower hypocretin-1 levels in patients with chronic and episodic cluster headache compared with controls (05).

Investigators searching for a relationship between pineal function in chronic migraine, cluster headache, chronic tension-type headache patients, and controls, found abnormal plasma melatonin curves in chronic tension headache and in patients with cluster migraine (10). Also, plasma melatonin values in patients with cluster headache during the cluster period showed an abnormal pattern.

Sleep deprivation may lower the threshold for pain and increase the risk of headaches. In a study, 24 participants were subjected to 3 days of sleep deprivation (42). Spontaneous pain, including headache, muscle pain, stomach pain, generalized body pain, and physical discomfort significantly increased by 5 to 14 units on a 100-unit scale during total sleep deprivation, compared to the normal sleep condition. Urinary prostaglandin-E2 metabolite significantly increased by about 30% in total sleep deprivation compared to normal sleep condition. Total sleep deprivation-induced increase in spontaneous pain, in particular headache and muscle pain, was significantly correlated with increase in prostaglandin-E2 metabolite. The authors of the study concluded that activation of the prostaglandin-E2 system is a potential mediator of increased spontaneous pain in response to insufficient sleep.

Epidemiology

Patients seen in headache clinics report a high prevalence of nocturnal or sleep-related headaches. In a study, 17% of the total headache group reported headaches occurring during the nocturnal or early morning (before final awakening) period. Fifty-five percent of patients in the headache subgroup had a specific sleep disorder identifiable by nocturnal monitoring in a sleep center (74). Approximately 20% of migraineurs had excessive daytime somnolence in a population-based sample (54). Cluster headache is predominantly a nocturnal disorder: 75% of cluster headaches appear between 9 PM and 10 AM, and half of them are associated with REM sleep (78). Cluster headache is associated with a sleep apnea syndrome only in the active cluster episode. Only one patient of five showed benefit with respect to cluster headache attack frequency (29).

The prevalence of sleep apnea headache in patients with obstructive sleep apnea is 12% to 18%, which is higher than morning headache with similar symptomatology in the general population (5% to 8%). Oxygen desaturation alone cannot explain the pathophysiology of sleep apnea headache, and the cause of sleep apnea headache remains to be elucidated (88). In a study, headache frequency was unrelated to severity of obstructive sleep apnea syndrome, albeit patients with sleep apnea had more headaches than control subjects (02). In another study of 4759 patients diagnosed with obstructive sleep apnea, the prevalence of tension-type headache was 10.2% among patients with sleep apnea and 7.7% among non-apnea patients (p <  0.001). The results suggest that patients with obstructive sleep apnea have a higher likelihood of developing tension-type headache than patients without sleep apnea (16).

Children with migraine may have an increased incidence of disturbed sleep and parasomnias, somnambulism, night terrors, and enuresis. Prolonged deep sleep is a risk factor for the provocation of sleep terrors and somnambulism, as well as for triggering migraine attacks in susceptible patients; these observations may have pathogenetic relevance. Ninety percent of the population grinds their teeth, but only in 5% will bruxism present as a clinical condition. In a study of 887 school-aged children in Spain, the authors found that the population with headaches had a greater number of sleep disorders, mainly insomnia, excessive daytime sleepiness, and parasomnias (98). The authors concluded that it is important to gather information about sleep habits in any child who consults for headache. Interestingly, the prevalence of pain, sleep problems, and fatigue, in addition to headache in boys and girls, increased significantly (p< .05) from 1989 to 2005 in Finland (64). All symptoms were associated with each other and with the child's psychiatric problems. However, psychiatric problems did not explain the observed increase in the symptom frequencies. Cerminara and colleagues reported three children between 7 and 11 years old with features of hypnic headache (14). These cases may expand the clinical spectrum of hypnic headache.

Patients with narcolepsy reported an abundance of headaches in a survey conducted in Germany (18). A group of 68 patients with idiopathic narcolepsy were interviewed for the presence of headaches. Eighty-one percent reported headaches that fit an International Headache Society diagnosis whereas 54% (64% women, 35% men) had migraine with the full complement of International Headache Society criteria.

In a study of 56 patients with obstructive sleep apnea syndrome compared to 50 patients with insomnia, the authors reported that in patients with obstructive sleep apnea syndrome, headache had most frequently a tension-type pattern and occurred on awakening in 74% of patients (01). Morning headache was common in habitual snorers and associated with poor health-related quality of life (15).

Insomnia is the most common sleep complaint among patients with chronic migraine (104). In another study the prevalence of sleep disorders was evaluated in 105 consecutive patients with daily or nearly daily headache and in 102 patients with episodic headache, matched by age, sex, and type of headache at onset (91). Patients with chronic headache showed a high prevalence of insomnia, daytime sleepiness, and snoring with respect to controls (67.7% vs. 39.2%, 36.2% vs. 23.5%, and 48.6% vs. 37.2%, respectively). Forty-five patients with chronic headache (42.9%) had psychiatric comorbidity (anxiety or depressive disorders) versus 27 patients with episodic headache (26.5%).

In a probability sample of adults in the United States, a structured interview administered by trained interviewers was used (57). Diagnostic criteria for migraine and migraine with aura were based on the International Headache Society classification. The presence or absence of four forms of sleep disturbance associated with an insomnia diagnosis was ascertained. The authors found a significant association between frequent severe headache, including migraine with and without aura, and disordered sleep. Adults with headache reported more frequent difficulty initiating sleep (odds ratio [confidence interval] = 2.0 [1.6-2.5]), difficulty staying asleep (2.5 [2.1-3]), early morning awakening (2.0 [1.7-2.5]), and daytime fatigue (2.6 [2.2-3.2]) compared to the individuals without headache. The authors concluded that adults with severe headache are at a significantly higher risk of suffering from sleep disorders, regardless of specific headache type. Optimal treatment of headache must include investigation for sleep disorders and vice versa.

Prevention

Proper sleep hygiene is paramount to aid in the prevention of sleep-related headaches. Bruni and colleagues evaluated the effect of modifying bad sleep habits across several headache parameters in 70 migraineurs with poor sleep hygiene (11). They instructed half of the group to improve their sleep hygiene while giving no directions to the other half. Mean duration and frequency of migraine attacks were significantly reduced at follow-up in the group asked to maintain proper sleep hygiene. The authors concluded that treatment through a simple modification of sleep behavior is an alternative approach to the treatment of migraine. Daily administration of preventive therapy is considered when migraine attacks occur more than twice a month or when they are prolonged and refractory to acute therapy. All effective headache medications interact with the serotonergic system and, thus, will likely have some corollary effect on sleep. Cluster headaches are prevented with avoidance of triggering factors, foremost of which is alcohol consumption. Sleeping late in the morning has been cited as a precipitating factor that should be avoided by patients with cluster headache.

Successful treatment of sleep apnea syndrome with nasal CPAP, bilevel positive airway pressure, or tracheostomy will eliminate associated headaches on awakening.

Episodes of bruxism may be reduced with stress management and are eliminated with diazepam, although the latter may not be a therapeutic option. In a 12-week randomized, double-blind, placebo-controlled, parallel-group study, 60 patients with migraine without aura were randomized to receive memantine (10 mg/day) or placebo for 12 weeks (69). The primary outcome was the monthly attack frequency at week 12 between the two groups. The authors found that memantine might be a tolerable and efficacious option for prophylaxis in patients with migraine without aura.

In a study of noninvasive vagus nerve stimulation (nVNS), 20 patients with treatment-refractory migraine were enrolled in a 3-month, open-label, prospective observational study (55). The patients received nVNS prophylactically twice daily and acutely as adjunctive therapy for migraine attacks. Outcomes were pain intensity, number of headache days per month, number of migraine attacks per month, number of acutely treated attacks, time to achieve pain relief, and sleep quality. Treatment with nVNS was safe and provided decreases in the frequency, intensity, and duration of migraine attacks. Improvements in sleep quality were also noted.

Differential diagnosis

Each headache evaluation should include at least a brief sleep history. Migraine and cluster headaches occurring at night need to be differentiated from other acute severe headaches such as those associated with intracranial brain tumors, ruptured aneurysm, and meningitis. Patients with intracranial tumors may be awakened at night by headache that improves on getting out of bed. Headaches on awakening, as observed in sleep apnea patients, are also seen in patients with severe hypertension, depression, intracranial tumor, muscle contraction headache, alcohol intoxication, and craniofacial sinus disease. Causes for concern are first or worst ever headache, associated neurologic symptoms or signs, progressive worsening of headache over days or weeks, intractable nausea or vomiting, fever, lethargy, confusion, and stiff neck. Hypnic headaches differ from cluster headaches and chronic paroxysmal hemicrania by their lack of unilateral pain with autonomic symptoms and their older age of onset.

Diagnostic workup

Neurologic consultation, neuroimaging studies, and lumbar puncture are indicated in patients who exhibit causes for concern. Nocturnal polysomnography is indicated for the study of patients suspected of having sleep apnea disorder or recurrent parasomnias. Videotaping should always be included in the polysomnographic study of parasomnias. Somnambulism and sleep terrors are associated with stage N3 of sleep. Patients with migraine, cluster headache, and hypnic headache may wake up with an acute attack more frequently during stage REM than during other stages of sleep, and those with cluster headache may suffer the attack at the same time of the night every night. Nocturnal polysomnography in patients with hypnic headache has found no consistent results (24). Attacks of chronic paroxysmal hemicrania may be so closely linked to REM sleep that they have been termed "REM sleep-locked." Bruxism occurs in stage N2 of sleep and less commonly in stage REM. The exploding head syndrome appears during nocturnal awakenings.

The use of polysomnography has been recommended in patients complaining of morning and nocturnal headaches (73). In a study of 25 patients with headache, Paiva and colleagues found 21 with disturbed sleep. The clinical diagnosis was reassessed after polysomnography in 13 patients due to the finding of obstructive sleep apnea, periodic limb movements, alpha-delta sleep, and insomnia. In a polysomnographic study of 30 migraineurs without aura, Karthik and colleagues found significantly lower sleep efficiency, prolonged sleep-onset latency, reduced NREM sleep, and more awakenings in migraineurs compared to controls (50). The authors were unable to detect whether these abnormalities were due to migraine or associated comorbidities.

Clinical evidence strongly supports screening for sleep disorders by headache practitioners. Treating sleep disorders may improve or resolve headache (84).

Management

Medications used to treat headache disorders may have a considerable impact on sleep physiology (68). Patients with migraine, cluster headaches, and chronic paroxysmal hemicrania should be instructed in good sleep hygiene and should avoid such potential precipitating factors as sleep deprivation, excessive sleep, stress, trauma, and ingestion of certain idiosyncratic foods including alcohol. Preventive treatment of migraine includes beta-blockers; calcium channel blockers; serotonin receptor antagonists (methysergide, only for use in periods not to exceed 4 weeks) and 5-HT2 antagonists cyproheptadine and methylergonovine; antidepressants that interact with serotonergic receptors such as tricyclics, MAO inhibitors, and serotonin reuptake inhibitors (fluoxetine and sertraline); anticonvulsants, particularly in children with abnormal EEG; and nonsteroidal anti-inflammatory agents. Sumatriptan, a 5-HT1 selective agonist, administered via subcutaneous injection (6 mg, may repeat after 1 hour, limit two injections in 24 hours) is an effective abortive medication for migraine attacks. Other abortive medications include ergotamine derivatives, acetaminophen, corticosteroids, and nonsteroidal anti-inflammatory derivatives. Symptomatic treatment for migraine attacks includes nonsteroidal anti-inflammatory derivatives, mixed barbiturate and analgesics, antiemetics (promethazine 50 mg), and in special circumstances of severity meperidine (50 mg) or codeine sulfate (30 mg). A report of a 2006 European Federation of Neurological Societies task force for the drug treatment of migraine recommends oral nonsteroidal anti-inflammatory drugs and triptans for the acute treatment of migraine attacks (28). The administration should follow the concept of stratified treatment. Before intake of nonsteroidal anti-inflammatory drugs and triptans, oral metoclopramide or domperidone is recommended. In very severe attacks, intravenous acetylsalicylic acid or subcutaneous sumatriptan are recommended as first choice drugs. Status migrainosus can be treated with steroids. Prophylaxis of migraine may be exercised with beta blockers (propranolol and metoprolol), flunarizine, valproic acid, and topiramate as drugs of first choice. Second choice drugs for migraine prophylaxis are amitriptyline, naproxen, and bisoprolol. Observational studies support the efficacy of melatonin in migraine prophylaxis (60). Immediate-release melatonin 3  mg, a melatonin receptor agonist (Agomelatine) 25  mg, and prolonged-release melatonin 4  mg were efficacious. Melatonin was effective after administration for 3 months. A word of caution was added as rare severe adverse effects have been noted, including liver disease, reproductive system dysfunctions, and decreased immunostimulation. The authors conclude that melatonin is a promising alternative for migraine prophylaxis. Cluster headaches are prevented with ergotamine derivatives at bedtime (1 to 3 mg sublingual), amitriptyline (150 mg daily), methysergide (6 to 8 mg daily), prednisone (40 mg daily), and lithium carbonate (initial dose 250 mg). An acute attack may be terminated with inhalation of oxygen.

In one triple-blind study of 68 patients with migraines treated with atorvastatin or placebo, in combination with nortriptyline, patients in the atorvastatin group experienced significantly fewer migraine attacks than the placebo group (P-value = 0.004) at 24 weeks (94). The most common adverse effects in intervention and control groups were constipation and insomnia, respectively. The authors concluded that statins seem to be a potential promising drug for prophylaxis of migraine headaches.

Chronic paroxysmal hemicrania responds specifically to indomethacin (50 mg at bedtime or 25 mg three times a day). Morning headaches related to sleep apnea syndrome disappear with the effective management of sleep apnea. Dodick and colleagues reported successful prophylaxis of hypnic headaches by drinking coffee at bedtime (two patients); administration of ergotamine tartrate 0.6 mg, phenobarbital 40 mg, and belladonna 0.2 mg at bedtime (nine patients); atenolol 25 mg at bedtime (16 patients); and aspirin 325 mg and caffeine 40 mg at bedtime (18 patients) (26). Indomethacin 25 mg at bedtime relieved symptoms in one case (46). One report reiterates the successful elimination of hypnic headache with the administration of indomethacin 150 mg daily over a period of 1 month (12). Moderate doses of lamotrigine have been reported as an alternative favorable treatment for hypnic headache (71). Successful prophylaxis with lithium carbonate has also been reported (25). There is a case report of a topiramate-responsive hypnic headache (53).

Reassurance and administration of clomipramine are curative in most instances of exploding head syndrome. Topiramate may be an alternative method for reducing the intensity of events in exploding head syndrome (75).

Bruxism is treated with stress management, a mouth guard, or an intraoral occlusal splint. For short-term management, diazepam (5 mg) given at bedtime will reduce the tooth-grinding activity.

The symptoms of post-concussional syndrome may respond to the administration of tricyclic antidepressants. Good to excellent relief of symptoms has also been obtained with the administration of intravenous dihydroergotamine and metoclopramide. Drug-related headaches tend to be benign and respond to conventional analgesics or disappear gradually with continued use of the drug. If necessary, discontinuation of the offending drug should be recommended, an action generally followed by disappearance of the headache.

Headache patients should be evaluated for the presence of obstructive sleep apnea as treating obstructive sleep apnea may improve headaches in some patients. In a retrospective chart review of all patients referred to an adult neurology clinic for headaches and sent for polysomnography in the span of 1 year, 82 headache patients were studied (48). Females predominated (70 females, 12 males), and the mean age was 45 ± 13 years with a mean body mass index of 32 ± 9. Fifty-two patients (63%) had obstructive sleep apnea. Increasing age, female gender, and chronic migraine without aura were predictive of obstructive sleep apnea. Patients with obstructive sleep apnea who were CPAP adherent (21/27) were more likely to have improvement in headaches than patients intolerant of CPAP (2/6), those who did not use CPAP (8/19), and those who did not have obstructive sleep apnea (16/30) (p = .045). Of the 33 patients who used CPAP, 13 reported improvement in headaches specifically due to CPAP therapy, and 10 additional patients noted benefit in sleep symptoms.

Severely obese migraineurs experienced marked alleviation of headaches after significant weight reduction following bariatric surgery. In a prospective observational study, 24 patients with migraine headaches were assessed before and 6 months after bariatric surgery (07). Patients were mostly female (88%), middle-aged (mean age 39.3), and severely obese (mean body mass index 46.6) at baseline. The mean number of headache days was reduced from 11.1 ± 10.3 preoperatively to 6.7 ± 8.2 postoperatively (p < 0.05) after a mean weight loss of 49.4%. Reductions in severity were also observed (p < 0.05); the number of patients reporting moderate to severe disability decreased from 12 (50.0%) before surgery to 3 (12.5%) after surgery (p < 0.01).

Migraine patients are commonly affected by sleep complaints, and the ketogenic diet is considered an option for the treatment of migraine. A study of ketogenic diet for the prevention of migraine included 70 migraine patients (65). The authors gathered information about subjective sleep complaints, such as insomnia, sleep quality (Pittsburgh Sleep Quality Index [PSQI]), and excessive daytime sleepiness by the Epworth Sleepiness Scale (ESS). They found that after 3 months of ketogenic diet therapy, migraine significantly improved, along with insomnia symptoms and excessive daytime somnolence. The authors concluded that the ketogenic diet may improve sleep complaints in migraine patients, independently of migraine improvements and anthropometric changes.

A single case presentation reports alleviation of exploding head syndrome (newly renamed as episodic cranial sensory shock) in a patient with a sleep disturbance that responded to daily single-pulse transcranial magnetic stimulation (sTMS) (81). Following treatment, the patient reported a significant reduction in the episodes of exploding head syndrome but not of concomitant migraine.

Special considerations

Cannabis is increasingly used to treat sleep disorders. In a systematic review of 568 studies, 31 were included for analysis to examine the effects of cannabis on sleep (99). Adverse events included headaches, along with sedation and dizziness. These adverse events occurred more frequently at higher doses. The authors concluded that high-quality evidence to support cannabis use for sleep remains limited.

Pregnancy

Remissions of migraine attacks and cluster headaches are common during pregnancy. Preventive medications should be discontinued during pregnancy. Ergotamine and its derivatives are to be avoided during pregnancy. On the other hand, migraine headaches and sleep disturbances are common comorbid conditions among pregnant women as noted in a cross-sectional study conducted among 1324 pregnant women. The odds of sleep disturbances were particularly elevated among pre-pregnancy overweight migraineurs (82). Memantine might be advantageous over other antimigraine medications in pregnancy because of tolerability, short duration required for titration, and safety profile (69).