Different sleep onset criteria at the multiple sleep latency test (MSLT): an additional marker to differentiate central nervous system (CNS) hypersomnias.

Excessive daytime sleepiness (EDS) has different correlates in non-rapid eye movement (NREM) [idiopathic hypersomnia (IH) without long sleep time] and REM sleep [narcolepsy without cataplexy (NwoC) and narcolepsy with cataplexy (NC)]-related hypersomnias of central origin. We analysed sleep onset characteristics at the multiple sleep latency test (MSLT) applying simultaneously two sleep onset criteria in 44 NC, seven NwoC and 16 IH consecutive patients referred for subjective EDS complaint. Sleep latency (SL) at MSLT was assessed both as the time elapsed to the occurrence of a single epoch of sleep Stage 1 NREM (SL) and of unequivocal sleep [three sleep Stage 1 NREM epochs or any other sleep stage epoch, sustained SL (SusSL)]. Idiopathic hypersomnia patients showed significantly (P<0.0001) longer SusSL than SL (7.7±2.5 versus 5.6±1.3 min, respectively) compared to NwoC (5.8±2.5 versus 5.3±2.2 min) and NC patients (4.1±3 versus 3.9±3 min). A mean difference threshold between SusSL and SL ≥27 s reached a diagnostic value to discriminate IH versus NC and NwoC sufferers (sensitivity 88%; specificity 82%). Moreover, NC patients showed better subjective sleepiness perception than NwoC and IH cases in the comparison between naps with or without sleep occurrence. Simultaneous application of the two widely used sleep onset criteria differentiates IH further from NC and NwoC patients: IH fluctuate through a wake-Stage 1 NREM sleep state before the onset of sustained sleep, while NC and NwoC shift abruptly into a sustained sleep. The combination of SusSL and SL determination at MSLT should be tested as an additional objective differential criterion for EDS disorders.