In 1988, Raskin 1 described six older patients who for years awakened from sleep once or twice a night with bilateral headaches lasting 30 to 60 minutes. In half the patients, the headaches were associated with nausea but with no other autonomic symptoms. Headache episodes responded well to lithium carbonate. This benign headache disorder was termed hypnic headache syndrome. Since then, a number of additional patients have been reported. 2-6 However, the possible relationship of hypnic headache to the sleep cycle is not well known. We report a patient with hypnic headache starting at stage 3 slow wave sleep, as documented by polysomnography.
We describe the polysomnographic data of a 79-year-old woman with an 11-year history of nocturnal headaches that were clinically consistent with hypnic headache. A polysomnographic study showed arousal at stage 3 slow wave sleep because of a headache episode. Although this finding could be nonspecific, it suggests the possible relationship between stage 3 slow wave sleep and hypnic headache.
A 79-year-old woman presented with an 11-year history of nocturnal headaches. The headaches were usually of throbbing quality and diffuse in location, awakened her from sleep once each night, and lasted 30 minutes after taking an ergot derivative. No ptosis, lacrimation, rhinorrhea, nausea, or other symptoms were reported. General and neurological examinations were unremarkable. Blood cell count, erythrocyte sedimentation rate, standard biochemistry, and magnetic resonance imaging were all normal. A polysomnographic study showed a sleep architecture characterized by low efficiency (60%), an increase in sleep latency and wakefulness after sleep onset, an increase in percentage of stage 1 and slow wave sleep, a marked decrease in percentage of rapid eye movement (REM) sleep, and a normal percentage of stage 2 sleep. The apnea/hypopnea index was less than 5, and SaO2 was between 95% and 100% (average value for the entire night 97%). After several minutes in stage 2, the patient awoke with a headache episode with normal breathing during the first epoch of stage 3 slow wave sleep. Previous therapeutic trials with propranolol, methysergide, cyproheptadine, pizotifen, amitriptyline, fluoxetine, moclobemide, indomethacin, ibuprofen, naproxen, and venlafaxine (the patient was being treated with venlafaxine when polysomnography was performed) had been unsuccessful. Therapy with lithium carbonate 600 mg per day at bedtime did not relieve the headache episodes, but these improved dramatically with gabapentin, 1200 mg per day.
Our patient had a typical hypnic headache syndrome similar to that described by Raskin. 1 Because some of his patients reported dreams upon awakening with headache and because of the response to lithium, this author suggested that these headaches were associated with REM sleep. A relationship between REM sleep stages and other headache syndromes, such as nocturnal migraine, tension-type headache, cluster headache, and chronic paroxysmal hemicrania, has been suggested. 7
The reasons for the onset of hypnic headache during stage 3 slow wave sleep in our patient are uncertain. The discharge rate of serotonin-containing neurons in awake animals decreases to 50% as the animal falls into the slow-wave sleep stage and to 10% in the REM stage. 8 However, during the slow wave sleep stage there is also an increase in gamma aminobutryic acid activity 9 that should hypothetically protect against the development of headache. The changes in sleep architecture could be explained based on the treatment she was taking (venlafaxine 75 mg, twice a day) and to the analgesic she took after having the headache. The lack of a temporal relationship with REM sleep could also be explained by the fact that antidepressants provoke the appearance of ambiguous polygraphic patterns (ie, mixture of REM and non-REM sleep).
To our knowledge, polysomnographic studies have not been performed previously in patients with hypnic headache, and our patient awakened from stage 3 slow wave sleep because of headache. Although this finding could be nonspecific, the routine use of polysomnography should be a useful tool to clarify the relationship of hypnic headache to the sleep cycle.