How much prior sleep is adequate for the multiple sleep latency test?

620 T he Multiple Sleep Latency Test (MSLT) is an important tool in the evaluation of excessive daytime sleepiness. It is indicated for the diagnosis of narcolepsy and the evaluation of idiopathic hypersomnia. 1 However, the MSLT results can be affected by a variety of extraneous variables that must be controlled or minimized to obtain interpretable diagnostic data. 2 Prior sleep deprivation is the most worrisome and difficult to monitor extraneous variable. The article " Nightly Sleep Duration in the Two-Week Period Preceding Multiple Sleep Latency Testing " in this edition, examines this issue by comparing sleep duration from self-reported average nightly sleep time, sleep logs, and actigraphy. Results showed that subjective estimates of sleep time were longer than the time measured by actigraphy. Moreover, actigraphy was the only measure that showed a low but significant correlation with MSLT results. It was suggested that actigraphy might be a better way of measuring sleep duration before the MSLT. This article touches on several basic questions surrounding prior sleep duration and the MSLT. The first question is when should an MSLT be performed? It is common for sleep duration to vary across the week with partial sleep deprivation on work nights followed by " catch up " on the weekend. This pattern was evident in the participants in this study, who also reported about 4 naps (range 0-14) per week. Not surprisingly, the patients slept the longest (7.4 hr) on the polysomnogram (PSG) night when given the opportunity. Having this information, a clinician should consider a diagnosis of insufficient sleep and recommend increasing sleep time to eliminate excessive daytime sleepiness before performing an MSLT. If this is not effective, an MSLT would be warranted if the history could support a diagnosis of narcolepsy. In practice, however, estimates of sleep durations in the week prior to MSLT are not usually obtained, so the clinician is not aware of the potential role of sleep deprivation. This study dramatically demonstrates what clinicians could be missing. Although military personnel with truncated sleep opportunities are not typical patients, it is well established that most Americans are sleep deprived. 3 The next question is how should prior sleep time be measured? This study compared actigraphy, a single subjective estimate of sleep duration, and sleep time from measured line lengths on sleep logs. Actigraphy showed the shortest total sleep time (0.5-1.5 hr less) in the 2 weeks prior to …