OBJECTIVE: To compare diurnal and nocturnal electrophysiological data from narcolepsy patients with undetectable (<20 pg/mL), low (20-110 pg/mL) and normal (>110 pg/mL) cerebrospinal fluid (CSF) hypocretin-1 levels.

PATIENTS/METHODS: A total of 109 narcolepsy patients and 37 controls were studied; all had available CSF hypocretin-1 measurements. The sleep laboratory studies were conducted between 2008 and 2014. The study retrospectively examined measurements of sleep stage transitions in diurnal and nocturnal continuous polysomnography. The percentage distribution of time awake and rapid eye movement (REM) sleep, and the occurrence of sleep onset REM (SOREM) in the nocturnal polysomnography were also measured.

RESULTS: Participants with undetectable hypocretin-1 levels had significantly higher frequencies of transitions than controls and those with normal hypocretin-1 levels. Participants with low hypocretin-1 levels showed more transitions than controls and, in some cases, also more than those with normal hypocretin-1. Participants with normal hypocretin-1 failed to show any significant difference from the controls, except in the overall diurnal transitions.

CONCLUSION: Undetectable hypocretin-1 levels in particular, but also low hypocretin-1 levels, were associated with a less stable phenotype featuring more sleep state transitions and SOREM episodes. In addition, there was a distinction between nocturnal and diurnal REM sleep in hypocretin-deficient participants, expressed as increased diurnal REM sleep, which was not reflected in nocturnal sleep.