BACKGROUND AND PURPOSE: Cluster headache (CH) is a primary headache disorder characterized by severe attacks of unilateral pain following a chronobiological pattern. There is a close connection with sleep as most attacks occur during sleep. Hypothalamic involvement and a particular association with rapid eye movement (REM) sleep have been suggested. Sleep in a large, well-characterized population of CH patients was investigated.

METHODS: Polysomnography (PSG) was performed on two nights in 40 CH patients during active bout and one night in 25 age, sex and body mass index matched controls in hospital. Macrostructure and other features of sleep were analyzed and related to phenotype. Clinical headache characterization was obtained by semi-structured interview.

RESULTS: Ninety-nine nights of PSG were analyzed. Findings included a reduced percentage of REM sleep (17.3% vs. 23.0%, P = 0.0037), longer REM latency (2.0 vs. 1.2 h, P = 0.0012) and fewer arousals (7.34 vs. 14.1, P = 0.003) in CH patients. There was no difference in prevalence of sleep apnea between patients (38%) and matched controls (32%, P = 0.64) although the apnea index in patients was numerically higher (mean apnea-hypopnea index 10.75 vs. 4.93). No temporal association between nocturnal attacks (n = 45) and particular sleep stages was observed.

CONCLUSIONS: To date, this is the largest study of sleep in CH. It is demonstrated that REM sleep is affected which is in line with our current understanding of CH and hypothalamic involvement in the regulation of this sleep stage. Further, fewer arousals were found in CH patients but no association between apnea events or specific sleep stages. The findings support a central role of the hypothalamus and arousal systems in CH.