Multiple sleep latency measures in narcolepsy and behaviourally induced insufficient sleep syndrome

Abstract

Background Short mean latencies to the first epoch of non-rapid eye movement sleep stage 1 (NREM1) and the presence of ⩾2 sleep onset REM (SOREM) periods on multiple sleep latency test (MSLT) occur in both narcolepsy–cataplexy (NC) and behaviourally induced insufficient sleep syndrome (BIISS). It is not known whether specific MSLT findings help differentiate the two disorders. Methods We analyzed MSLT data including sleep latencies to and between different sleep stages of 60 age-, gender- and body mass index (BMI)-matched subjects (hypocretin-deficient NC, actigraphy-confirmed BIISS, healthy controls: each 20). Results Mean latency (in minutes) to NREM1 sleep was significantly shorter in NC (1.8 ± 1.5) than in BIISS (4.7 ± 2.1, p < 0.001) and controls (11.4 ± 3.3, p < 0.001). Mean latency to NREM2 sleep was similar in NC (8.6 ± 4.7) and BIISS (8.1 ± 2.7, p = 0.64); latency to either NREM2 or rapid eye movement (REM) sleep (i.e., the sum of the sleep latency to NREM1 and the duration of the first NREM1 sleep sequence), however, was shorter in NC (4.4 ± 2.9) than in BIISS (7.9 ± 3.5, p < 0.001). Referring to all naps with SOREM periods, the sequence NREM1–REM–NREM2 was more common (71%) in NC than in BIISS (15%, p < 0.001), reflecting the shorter latency from NREM1 to NREM2 in BIISS (3.7 ± 2.5) than in NC (6.1 ± 5.9, p < 0.001). Conclusions Our findings show that both sleepiness (as measured by NREM1 sleep latency) and REM sleep propensity are higher in NC than in BIISS. Furthermore, our finding of frequent REM sleep prior to NREM2 sleep in NC is in line with the recent assumption of an insufficient NREM sleep intensity in NC. Together with detailed clinical interviews, sleep logs, actigraphy, and nocturnal polysomnography, mean sleep latencies to NREM1 ⩽2.5 min, the presence of multiple SOREM periods, and the sequence NREM1–REM–NREM2 may be the best MSLT measures to discriminate NC from BIISS.