In a 2015 episode of The Simpsons, Homer is diagnosed with narcolepsy. Overwhelming sleepiness at the nuclear power plant lands him in the hospital. Sampling his spinal fluid (ouch!), Homer’s chuckling, deep-voiced doctor quickly performs a test for hypocretin, a brain chemical important for staying awake and regulating REM sleep.
Reality check: testing for hypocretin takes time, and is not currently available in the United States. Let’s talk about how sleep disorders such as narcolepsy and idiopathic hypersomnia are actually diagnosed: operationally, rather than biologically. The less flashy, but standard, way to assess patients is to ask them to take a series of five naps and see how fast they doze off, and how fast they go into REM sleep (the rapid eye movement dreaming phase).
This process, known as the Multiple Sleep Latency Test or MSLT, works pretty well for narcolepsy type 1, the more distinctive form of narcolepsy that includes cataplexy. And it’s hard to fake being sleepy enough to zonk out within a few minutes. But it has a bunch of problems, and dissatisfaction with the MSLT has been developing among sleep specialists for the last several years.
At Emory, neurologists Lynn Marie Trotti and David Rye published an analysis of what I will call the “flip flop problem” in 2013, with others in the field following up more recently. The flip flop problem is: someone who takes the MSLT one day will frequently get another result if they take it again on a different day. Also, if conducted with a large group of people in a community, the MSLT tends to tag many of them as having a sleep disorder, because of shift work or chronic sleep deprivation. Based solely on the MSLT, 1 in 30 people might have “narcolepsy,” but narcolepsy is actually much less common (often estimated at 1/2000). On the other side of the coin, MSLT results can tell a doctor that patients’ complaints of sleepiness – their lived experiences — don’t fit any disease.
“If we rely on the MSLT, we’re missing people who have an important clinical issue,” Trotti said at a workshop at the June 2018 Sleep research meeting in Baltimore.
The idea behind the MSLT is to separate out the effects of the environment and determine someone’s “objective level” of sleepiness. However, Trotti pointed out that the situation set up by the MSLT doesn’t correspond to the problems that bring patients to see a sleep specialist: “Nobody comes to clinic and says: ‘When I try to fall asleep, I do it too quickly.’”
In Baltimore, several investigators outlined additional ways to assess patients’ sleepiness and diagnose their sleep disorders. Some of the approaches they described – in the workshop, and also in posters — represent potential supplements or even alternatives to the MSLT. They included:
*actigraphy – estimating, with a wearable device, how much time someone sleeps at home
*the Maintenance of Wakefulness Test (MWT), in which people are put in a dark room and asked to stay awake as long as they can
*tracking how long someone can sleep in a single overnight stretch
*adding measures of fluctuations in pupil size in darkness as a measure of drowsiness
*extracting detailed information from an overnight sleep test on how much time someone spends in REM and non-REM phases of sleep (sleep architecture)
Many of these tests are not new. Actigraphic estimation of sleep time in IH is already included in the ICSD-3, the most recent International Classification of Sleep Disorders. The MWT is used to gauge sleepiness in pilots and commercial drivers, as well as in clinical trials. Including additional measures could bring more sleepy people into the treatment tent, and also might provide clues to physiological mechanisms. Challenges for the field could be to agree on how to combine these diagnostic approaches, and also where to draw lines of “normal” vs “abnormal.”
At the workshop, French neurologist Isabelle Arnulf described how a “sleep as long as you can” protocol is used in sleep centers in France and Italy. David Plante, from the University of Wisconsin, summarized how his sleep center has been using measurements such as: total sleep time, the PVT (psychomotor vigilance test — a workhorse measure of reaction time), and pupillometry, which tracks fluctuations in pupil diameter. Plante cited pupillometry’s use, going back to the 1960s, to study sleepiness in narcolepsy. Through a combination of these measures, Plante said, his team has been able to establish “objective hypersomnolence” in more than twice as many patients than with the MSLT alone.
In similar work, sleep researchers from Münster, Germany recently published an analysis of attention and alertness tests in patients with narcolepsy and IH diagnoses, some of whom had normal MSLT results. They found that people with narcolepsy type 2 and IH tended to have problems with vigilance but normal results for selective attention, which was not the case for narcolepsy type 1.
“Critical flicker fusion,” which Emory sleep researchers have begun experimenting with, is a vision-oriented test of nervous system function. It monitors at what point a light that is blinking at an accelerating speed will appear completely fused or steady. This test has a background in the study of hepatic encephalopathy: liver disease’s effects on the brain. Although they are different, what critical flicker fusion and pupillometry have in common is that they are fast and oriented toward vision. And in contrast to when patients are asked to stay awake, like in the MWT, it is less obvious how effort plays a role in modulating the results.
For IH, the analyses of sleep architecture are intriguing, but more speculative. One of the main burdens experienced by IHers is waking up in the morning, but still feeling the need for sleep. Are they missing some element of healthy sleep? On the surface, the answer appears to be no. Several studies have shown that IHers still have both REM sleep and slow wave sleep, the deepest phase of non-REM sleep. A recent meta-analysis by Plante finds IHers tend to have more REM and less slow wave sleep than controls, although the differences aren’t huge. For reference, see this 1980s comparison of narcolepsy and “idiopathic CNS hypersomnia” from Stanford.
In IH, maybe there is some finer defect in the brain’s nocturnal oscillations? In the recent poster presentations (Alyssa Cairns from SleepMed, Makoto Honda from Tokyo, Logan Schneider from Stanford) there was not a consensus. For the positive identification of narcolepsy type 1, separately from the MSLT, it is also possible to use the observation that muscle tone is maintained during REM sleep (presentation from Kiran Maski at Boston Children’s).
One Response to Measuring sleepiness: alternatives to five naps