The slope of performance during the psychomotor vigilance task: an additional indicator in the assessment of hypersomnolence? Free

In the last decades, there has been an increasing interest in measuring impaired cognitive performance in hypersomnolence disorders to improve diagnosis and characterize the phenotype of different conditions [1–3]. Disturbed vigilance and attentional problems are in fact one of the expressions of the excessive daytime sleepiness (EDS). They are seen across all central hypersomnia diagnoses [4–7] and may result in a significant deterioration of the patient’s quality of life or a risk to the patient’s safety, if occurring while, for instance, driving a car [8]. However, vigilance and attentional impairment received much less attention when studying people with central hypersomnia with respect to the most widely known expression of the EDS as the increased tendency or need to fall asleep [9]. Current methods to objectively assess EDS are, thus, mainly based on the multiple sleep latency test (MSLT) and the maintenance of wakefulness test (MWT), which measure the tendency to fall asleep and the ability to stay awake, respectively [9, 10].

The sustained attention to response task and the psychomotor vigilance task (PVT), which both test the vigilance and sustained attention, have proved to be very sensitive to sleepiness-related performance impairment in hypersomnolence conditions [11–16]. The PVT is one of the most widely used tools to assess behavioral alertness and sustained attention by recording reaction times (RT) to visual (or auditory) stimuli that occur at random inter-stimulus intervals [17]. Baseline PVT performance was found to be poorer in patients with hypersomnolence conditions (i.e. narcolepsy type 1, narcolepsy type 2, idiopathic hypersomnia, and also non-specified sleepiness, not due to central hypersomnia or insufficient sleep) compared to control cases [11, 14]. In particular, hypersomnolent patients performed significantly worse in terms of mean and median RT, response speed (1/RT), lapses (RT ≥ 500 milliseconds), fast and slow response speed (1/RT of the fastest and slowest 10% of responses), and recorded a higher time-on-task worsening of the response speed (i.e. the slope of the reciprocal RT during the test).

While the ability to fall asleep and stay awake depends on a range of physical and environmental factors, also the patient's personal motivation and cooperation play a role [18–20]. In this light, studies have shown that participants are able to significantly increase sleep latency on the MSLT when required to be as wakeful as possible on the test, or also, to decrease sleep latency on the MWT when instructed to appear as sleepy as possible [18, 20]. There is a tangible risk that the patient is not being completely unbiased in his/her cooperation during the MSLT and the MWT or have non-genuine behavior since test results may imply severe consequences such as potential job loss, driving license suspension, or the diagnosis of a lifelong medical condition treated by potential drugs of abuse. Mariano et al. designed an interesting study aimed to determine to which extent PVT performance could be altered intentionally and if there are some PVT metrics that could help to differentiate patients with hypersomnolent from participants just pretending to be sleepy [21]. PVT measurements were collected from 16 healthy participants instructed to fake sleepiness and were compared to previously published data [14] from 33 healthy participants instructed to achieve their best possible performance and patients with the diagnosis of central hypersomnia (i.e. idiopathic hypersomnia, narcolepsy type 2, and narcolepsy type 1). Healthy participants that simulated sleepiness had dramatically worse (lapses, 1/RT, fastest and slowest 10% 1/RT) or similar (mean and median RT) performance on PVT than that found in hypersomnolence patients with impaired psychomotor vigilance. However, the slope of response speed over time showed a significantly higher time-on-task worsening for patients lning with hypersomnolence than for controls while the difference was not so evident between controls simulating sleepiness and those performing their best. Thus, as controls did not receive any instructions on how to pretend sleepiness, they excessively altered their performance on PVT, but they weren’t able or did not consider adjusting their reaction time along the test to simulate fatigue.

Among the several PVT performance metrics provided by the literature, the slope of speed response is not so commonly used. The number of lapses and the response speed (i.e. both slowest 10% 1/RT and mean 1/RT) are considered as more sensitive to sleep deprivation and alertness, unlike the most widely used mean and median RT [22]. However, according to Mariano et al., these PVT metrics may not be the best indicators to measure sleepiness because it is possible for participants to voluntarily decrease the response time, while the slope of speed response seems a more valid measure to discriminate patients who simulate or exaggerate their sleepiness. As already mentioned, there might be an actual intention to decrease latencies on the MWT. In this case, when there is a reasonable suspicion that the patient might have a genuine interest in not appearing capable to stay alert, the PVT with the slope of performance as the outcome could be useful in association with the MWT in the assessment of alertness for fitness to drive or in the evaluation of treatment efficacy.

Changes in PVT performance, such as the overall slowing of response times and the increase in the number of lapses, related to sleep-deprived induced sleepiness can increase as task duration increases [23]. This time-on-task effect may reflect a state of instability between the involuntary drive to fall asleep and the counteracting drive to sustain alertness that influences behavior during periods of sleep loss and leads to unstable sustained attention [24]. In this context, the slope of performance over time may allow us to better explore the impairment of vigilance and attention in terms of interaction between sleep-initiating and wake-maintaining systems in central disorders of hypersomnolence.

In conclusion, although additional studies are needed to confirm Mariano et al. results, the slope of PVT performance over time seems to have the potential to be proposed as an additional indicator in combination with the usual psychomotor response speed and lapses for evaluating performance impairment in hypersomnolence disorders.

Disclosure statement

I report no financial and non-financial disclosures relevant to the manuscript. I declare no conflicts of interest related to this article.

Study funding

none

References