REM-suppressing medications can alter the ability to detect REM sleep. Medications that can alter sleepiness (stimulants) or REM sleep should be withdrawn at least 15 days before testing. A sleep diary or actigraphy should document a regular and sufficient amount of sleep for at least 7 days before the study. The purpose of the PSG is to exclude other causes of excessive daytime sleepiness such as OSA and periodic limb movement disorder (PLMD). Overnight PSG during the patient's habitual sleep period should precede the MSLT. The MSLT is the standard objective test for the assessment of sleepiness and the diagnosis of narcolepsy 53–55 ( Table 24–5 see also Table 24–4). Berry MD, in Fundamentals of Sleep Medicine, 2012 Multiple Sleep Latency Test 46 No correlation was found between the mean sleep latencies on the first and second tests, with a change in diagnosis occurring in 42% of patients because of differences in the mean sleep latencies. 8,9,10,12 In addition, a recent study of 36 patients demonstrated poor test-retest reliability of the MSLT in a clinical population of patients with non-hypocretin-deficient central nervous system hypersomnias.

Regarding these limits, some patients may have mean sleep latencies longer than 8 or even 10 minutes. The second reason is the obligation to wake the patient in the morning to perform the MSLT, thus precluding the recording of prolonged nighttime sleep, which is a typical symptom of IH with long sleep time.

8,9 The first reason is the usual difficulty keeping the patient awake before the test and between sessions of the test. MSLT is of limited diagnostic value in IH patients with long sleep time. 10,12 Subjective awareness of sleep during naps is often higher than in patients with narcolepsy. 9,10,12 SOREMPs occur in 3% to 4% of naps, but never more than once. MSLT typically shows a mean sleep latency score of less than 8 minutes, and a few IH patients have mean sleep latencies of less than 5 minutes. Bassetti, in Principles and Practice of Sleep Medicine (Sixth Edition), 2017 Multiple Sleep Latency Test A SOREMP on the prior night polysomnogram further suggests a diagnosis of narcolepsy. If other sleep disorders have been excluded, this will help to support a diagnosis of narcolepsy. Most narcoleptic patients have at least two SOREMPs on a single MSLT. There is significant overlap in mean sleep latency between normal individuals and narcoleptic patients, and therefore mean sleep latency data on MSLT are nondiagnostic. Patients with narcolepsy usually have a mean sleep latency of 2–3 min on any given study, this could be much longer. The mean sleep latency for normal individuals is about 10 min (2–19 min ☒ standard deviations). Conversely, increased antecedent total sleep time will decrease the mean sleep latency. In the normal non-sleep-deprived person, a shorter total sleep time just prior to the MSLT will result in an increased mean sleep latency. The mean sleep latency is affected by such factors as the number of naps, the patient's age, and total sleep time during the preceding week. During an MSLT, a mean sleep latency ≤8 min with greater than or equal to two SOREM periods is consistent with a diagnosis of narcolepsy in the absence of confounders. According to the AASM, a sleep latency of <8 min is diagnostic of sleepiness. Pathologic sleepiness is defined as a mean sleep latency <5 min and this has been associated with impaired performance. Normal adult mean sleep latency is between 10 and 20 min. In other words, the sleepier one is, the quicker they fall asleep. The MSLT assumes that sleep latency decreases with greater physiologic sleepiness. Anderson, in Encyclopedia of Sleep, 2013 MSLT Interpretation

Mild narcolepsy without cataplexy