Dr. Lynn Marie Trotti answers questions about diagnostics relating to hypersomnia and concludes…
A – For narcolepsy type 1 (or narcolepsy with cataplexy), measurement of hypocretin levels in cerebrospinal fluid is the most specific, most definitive diagnostic test available. It is a true “gold-standard” in that it measures the problem that is causing the symptoms (that is, patients with narcolepsy type 1 have symptoms because their brain can no longer produce hypocretin). However, this requires a lumbar puncture, and not all sleep physicians perform this procedure. Furthermore, clinical testing of hypocretin levels is not commercially available. In contrast, the multiple sleep latency test (MSLT) is a widely available test. Over time, MSLT diagnostic cut-offs have been changed to improve the ability of the MSLT to correctly diagnose cases of narcolepsy with cataplexy, and so this is the most widely used test for narcolepsy.
Although diagnostic criteria for narcolepsy without cataplexy require a multiple sleep latency test and diagnostic criteria for IH can include results from a multiple sleep latency test, research suggests that the MSLT may not be ideal for diagnosing these conditions. However, an optimal diagnostic test to replace the MSLT has not yet been developed and validated.
Q – Why is an 8-minute cutoff of mean sleep latency used for diagnosing idiopathic hypersomnia on the MSLT?
A – The use of an 8-minute sleep latency was a consensus decision made by the authors of the American Academy of Sleep Medicine’s International Classification of Sleep Disorders (ICSD). In the 2001 revision to the original ICSD, a mean sleep latency less than 10 minutes was reported as typically present in patients with IH but was not a mandatory part of the diagnosis.
However, by the time of the 2005 publication of the second edition (ICSD-2), this had been shortened to 8 minutes (and the 8-minute cutoff remains as one potential diagnostic criteria for IH in the current ICSD-3). The authors of the ICSD-2 decided to use an 8-minute cutoff “to define sleepiness for diagnostic purposes” based on the fact that this cutoff appeared to be the best one for diagnosing narcolepsy. The underlying assumption they made was that an 8-minute cutoff should work well for IH diagnosis if it worked well for narcolepsy diagnosis.
At the same time, however, they cited data that the average mean sleep latency in IH patients was 6.2 minutes, with a standard deviation of 3.0 minutes. This implies that NOT all patients with IH have a mean sleep latency less than 8 minutes. The apparent discrepancy between these two ideas was not explicitly addressed in the ICSD-2, although the authors did state in the text accompanying the diagnostic criteria that the mean sleep latency is “usually” less than 8 minutes in IH, and that sometimes additional diagnostic testing would be needed.
This became more explicit in the ICSD-3, which now allows documentation of long sleep times to establish a diagnosis of IH in patients with a mean sleep latency greater than 8 minutes. Whether the revised criteria are fully optimized for the diagnosis of IH remains to be determined.