Impaired cognition in narcolepsy: clinical and neurobiological perspectives

Abstract

In addition to well-known symptoms such as sleepiness and cataplexy, many people with narcolepsy have impaired cognition, reporting inattention, poor memory, and other concerns. Unfortunately, research on cognition in narcolepsy has been limited. Strong evidence demonstrates difficulties with sustained attention, but evidence for executive dysfunction and impaired memory is mixed. Animal research provides some insights into how loss of the orexin neurons in narcolepsy type 1 may give rise to impaired cognition via dysfunction of the prefrontal cortex, and cholinergic and monoaminergic systems. This paper reviews some of these clinical and preclinical findings, provides a neurobiological framework to understand these deficits, and highlights some of the many key unanswered questions.

Statement of Significance.

Many people with narcolepsy are bothered by cognitive challenges such as difficulty sustaining attention. This paper reviews clinical and preclinical research on this topic and highlights important, unanswered questions.

When managing narcolepsy in the clinic, providers often focus on optimizing wakefulness and minimizing cataplexy, but, even on medications, people with narcolepsy often struggle with cognitive difficulties which adversely impact their lives. In a survey of 1699 people with narcolepsy type 1 (NT1) and type 2 (NT2), the most burdensome symptoms included cognitive difficulties, including difficulty thinking, remembering, concentrating, and paying attention [1]. In fact, cognitive difficulties were reported by half of all respondents and came in second only to excessive daytime sleepiness (EDS) as the most significant symptom affecting their daily lives. Despite this substantial impact, impaired cognition continues to be an underrecognized aspect of narcolepsy.

Recent reviews provide detailed information on the neuropsychology of narcolepsy [2–5], and here, we review this briefly with our major focus on the potential neurobiology of impaired cognition in narcolepsy and research ideas for understanding this better.

Neuropsychology of Narcolepsy

The cognitive difficulties of narcolepsy are generally assumed to be a consequence of EDS [6, 7], but even when people with narcolepsy are at their most alert, impaired cognition can be a frequent challenge (see Box 1). Patients often describe difficulties in maintaining attention and the need to exert much effort to do even simple tasks. Many people with narcolepsy report that their thinking is best after a period of sleep, and it often worsens across the day. Long, boring, and sedentary tasks such as studying, lengthy tests, and long drives are especially problematic; patients can have automatic behavior when drowsy in which they continue with some action such as driving or eating, but with little awareness, self-control, or memory of the event.

Box 1. Self-reported perspectives on cognition in narcolepsy.

People with narcolepsy report a variety of concerns about their cognition, and these fall into several broad domains.

• (1) Inattention

  • a. “When I was in grad school, I would read a chapter three times, but I still didn’t know what it said.”

  • b. “Even when I’m well-rested and taking all my meds, studying is harder than before I had narcolepsy.”

  • c. “My meds keep me from falling asleep at work, but I spend much of the day in a fog.”

• (2)Poor memory

  • a. “I have to double-check if I sent that email.”

  • b. “My memory is affected severely by my narcolepsy which sometimes gives off the impression that I don’t care or don’t pay attention, but I just can’t remember.”

• (2)Mental effort

  • a. “I’m exerting top–down effort in everything I do.”

  • b. “I can’t multitask because the task at hand takes so much effort.”

• (3)Automatic behavior

  • a. “I’ve driven miles past my usual exit off the highway and poured orange juice into my breakfast cereal.”

  • b. “The worst part about automatic behaviors is doing and saying things I might not have intentionally done and that damages my relationships.”

• (4)Effects of medications

  • a. “When I was taking Adderall, I couldn’t switch tasks and felt ‘rabbit-holed’—unable to leave things alone. Now that I’m off it, I feel less hyper-focused, and my creativity is better.”

  • b. “My meds affect the way I respond to things because my brain feels overworked sometimes from being awake, so my thinking is almost never entirely clear.”

Though people with narcolepsy often use the term “brain fog,” there is no formal definition, and the meanings can range from difficulty processing information to simple cognitive tasks taking much effort, to poor recall.

Subjective reports and objective tests are both useful though they have their limitations. Reports of patients and their families and the observations of their clinicians can shed light on the nature, frequency, and severity of cognitive difficulties, including the extent to which these impact participation in family life, education, and employment. However, these reports can be influenced by an individual’s mood, sense of control, level of health, level of education, language ability, trust with the clinician, and level of social support.

Objective assessments of cognition can characterize the nature and magnitude of these cognitive difficulties by measuring different aspects of cognition using standardized tests [8, 9]. The pattern of strengths and weaknesses across different cognitive domains provides a basis for understanding an individual’s cognitive impairment, as well as offering clues about the underlying neural circuitry [8–10]. However, performance on objective tests reveals cognitive abilities at the time of assessment and may not reflect difficulties experienced at other times of the day or in everyday situations [11, 12]. The validity of these tests also depends upon the participants’ understanding of the test requirements and adequate psychometric properties (i.e. no floor or ceiling effects). Additionally, complex cognitive processes depend upon simpler cognitive processes [8]. For example, executive function or memory may be impaired by deficits in attention or processing speed. As such, signs of impaired executive function or memory in a person with narcolepsy could be due to inattention or EDS.

Currently, neuropsychological research in narcolepsy is limited as studies have focused mainly on NT1 with very little work on NT2, have included relatively small numbers of participants (often <30), and have generally measured only one or two domains of cognition [12–18]. Still, these studies provide consistent signs of impaired attention with mixed evidence of impaired executive function and memory (for reviews see [2–5]). Limited data suggest that current medications may modestly improve attention and executive functioning, but the real-world impacts of medications on cognition in people with narcolepsy have not been clearly demonstrated [19–24]

Attention

Adequate attention allows one to focus awareness on specific stimuli, information, or tasks, while simultaneously filtering or rejecting less relevant information. Normal attention is essential for nearly all other aspects of higher cognition such as learning and memory, decision-making, and problem-solving.

People with narcolepsy often report difficulty engaging, focusing, and sustaining attention, and many studies have demonstrated impairments in sustained attention in adults and children with narcolepsy (briefly summarized in Supplementary Table 1) [12, 14, 15, 17, 18, 22, 25–28].

Sustained attention is often measured using the Psychomotor Vigilance Test (PVT) or the Sustained Attention to Response Task (SART) [7, 29, 30]. The PVT requires a participant to maintain attention over relatively long periods (typically 10 minutes) and rapidly press a button in response to an infrequent visual stimulus such as a small light or a spot on a screen [31]. The main measure of attentional impairment in the PVT is the number of events which fail to trigger responses (i.e. misses or lapses). The SART also requires rapid responses to visual stimuli, but in addition, participants must inhibit responses to one type of these stimuli; for example, participants are instructed to press a button when a number between 0 and 9 is presented on a screen except when it is a 3 [32]. The SART takes less than 5 minutes, and attentional impairment is defined by lapses (omissions) plus responses that are not inhibited (commissions or false-positive responses). Whereas both the PVT and the SART assess sustained attention, the SART also assesses response inhibition which is an aspect of executive functioning.

Several studies have shown worse function and more variability performance on the PVT in unmedicated people with NT1 and NT2 relative to healthy controls [24, 33–35]. Although the number of participants in each of these studies was relatively small, people with NT1 and NT2 consistently showed more lapses in the PVT than controls. Additionally, Trotti et al. [35] found that more than half of the patients with NT1 had lapses outside two standard deviations of the mean number of lapses in controls, suggesting that this worse performance meets the criteria for clinical impairment ([36]). Similarly, performance on the SART was also worse in people with NT1 than in healthy controls in several studies [24, 37, 38].

Only a few studies have examined other attentional processes such as selective attention and divided attention. Results from these studies are mixed, with some finding impaired performance in narcolepsy while others do not [12, 14–18, 25, 26, 28, 39]. Thus, impaired sustained attention is a common finding in narcolepsy, but further detailed studies are needed to develop more complete models of attentional challenges.

Executive function

Executive function is an umbrella term that includes cognitive processes such as problem-solving, forming strategies, and performing mental calculations, and impaired executive function can interfere with successfully performing instrumental activities of daily living. Some studies of adults with narcolepsy demonstrate impaired executive function with poor performance on tests of executive control, working memory, initiation, inhibition, verbal fluency, and problem-solving [13–15, 17, 40]. However, other studies report no differences between patients with narcolepsy and matched controls on tests of verbal fluency, set-shifting, problem-solving, and inhibition [15, 23, 39, 41]. Considered together, this research suggests that people with narcolepsy may have difficulties with executive function, but more research is required to understand the nature and magnitude of such impairment and whether it is a consequence of impaired attention.

Memory

Memory requires encoding information from the environment, storing that information, and then retrieving that information when needed. People with narcolepsy often voice concerns about their memory, but studies objectively measuring memory processes in narcolepsy have found inconsistent evidence of impairment [12, 16, 17, 26, 39, 42].

As with complaints about memory, self-reports of other cognitive difficulties (e.g. difficulty thinking fast) sometimes do not correlate with objective testing of those abilities [12, 22, 43–46]. This mismatch may reflect the different ways in which subjective and objective assessments capture information about cognition. First, people tend to describe their cognitive difficulties in terms of how it impacts their daily lives rather than the cognitive processes that give rise to them; for example, “I keep forgetting people’s names” rather than “I’m not paying attention when people introduce themselves.” Second, subjective reports of cognitive impairment generally reflect difficulties over weeks or months, and these can be influenced by recall or expectation bias. Third, performance on objective tests reflects cognitive abilities at the time of assessment and may not reflect the difficulties experienced in everyday situations [11, 12]. For example, a student with narcolepsy may be able to perform well on a neuropsychological test assessing her ability to recall a list of words administered in the clinic but may find it difficult to learn material for a test, which requires learning and integrating complex information over hours or days rather than learning and recalling a list of words over 20–30 minutes. Fourth, for some people, the effects of depression and anxiety may contribute to perceived cognitive difficulties [12, 13, 47]. Thus, one should weigh both subjective and objective sources of information about cognition to gain a fuller understanding of an individual’s capabilities.

Are these cognitive impairments simply due to sleepiness?

Some have proposed that sleepiness is the main driver of impaired cognition in narcolepsy [6, 13, 46]. In healthy individuals, sleep deprivation and sleepiness impair cognitive functioning [48, 49], and many people with narcolepsy feel their cognition is best after sleeping [47]. However, several studies have found that cognitive performance correlates only weakly or not at all with EDS [12, 17, 18, 22, 25, 37, 38, 40]. In addition, people with narcolepsy had impaired performance on a complex cognitive task, but healthy individuals deprived of sleep for 32 hours showed good performance even though they had a similar level of EDS [47]. Furthermore, people with NT1 generally show greater cognitive impairments than those with NT2 or idiopathic hypersomnia (IH), even though all have problematic EDS [14, 18]. Considered together, we feel EDS contributes to impaired cognition in narcolepsy, but specific effects of orexin deficiency are likely important as discussed in the next section.

Neurobiology of Impaired Cognition in Narcolepsy

NT1 is caused by severe loss of orexin-producing neurons, and this section reviews some of the key neural circuitry of cognition and how the loss of orexin signaling may impact these circuits. The cause of NT2 remains unknown, but it may be due to less extensive loss of these cells so some of these same mechanisms may apply [50].

The orexin-producing neurons in the lateral hypothalamus project throughout the brain, promoting wakefulness, regulating rapid eye movement (REM) sleep, and integrating motivation signals with arousal responses (Figure 1) [51]. The prepro-orexin gene produces two peptides, orexin-A, and orexin-B, which bind differently to the OX1R and OX2R receptors: orexin-A binds strongly to both receptors, while orexin-B has a higher affinity for OX2R. Functionally, OX1R appears to be important for reward and motivation responses, whereas OX2R is crucial for regulating wakefulness and REM sleep. Which orexin receptors are crucial for cognition remains unknown, but both are expressed by neurons crucial for cognition, including those in the prefrontal cortex, the basal forebrain (BF) cholinergic neurons, ventral tegmental area (VTA) dopaminergic neurons, and locus coeruleus (LC) noradrenergic neurons.

Figure 1.

Cognition relies on networks within the prefrontal cortex, which are modulated by ascending signals from noradrenergic neurons of the locus coeruleus, dopaminergic neurons of the ventral tegmental area, and cholinergic neurons of the basal forebrain. All these regions are excited by the orexin neurons, and loss of orexin signaling likely contributes to the cognitive problems of narcolepsy.

Prefrontal cortex

The prefrontal cortex (PFC) is essential for higher cognitive processes, including maintaining attention and executive functions such as decision-making, problem-solving, and adapting to new situations. In addition to connections with many brain regions elsewhere, the PFC contains functionally specialized networks (e.g. central executive, dorsal attention, and default mode networks) which support specific aspects of cognition. Brain connectivity studies in humans suggest that these PFC networks coordinate on complex tasks; when performing a complex task such as a working memory task, the connectivity between these networks strengthens, while distinctions between them diminish [52]. In fact, this coordination is necessary as extensive, bilateral lesions across the entire PFC of primates cause severe learning impairments, but lesions confined to subdivisions of the PFC do not markedly compromise specific cognitive tasks [53, 54].

The orexin neurons innervate the PFC and other areas of the cortex [55], and infusion of orexin-B in the PFC enhances performance on an attentional task in rats [56]. OX1R and OX2R mRNA can be seen in several cortical lamina [57], but orexins appear to directly excite only lamina 6b neurons that help suppress slow waves and enhance gamma activity [58, 59]. Gamma oscillations (30–80 Hz) can synchronize activity across brain networks and are likely crucial for many cognitive tasks including attention and working memory [60, 61]. Orexin-deficient mice have lower gamma oscillations [62], and an OX1R antagonist reduces EEG gamma wave activity [63]. These studies demonstrate that orexins may act directly in the PFC to improve cognition, but as detailed below, the orexin neurons also strongly excite many key subcortical systems which are essential for normal PFC function.

Cholinergic neurons of the BF

Cholinergic neurons in the BF enable normal attention and executive function through their projections to the PFC and elsewhere [64]. Chemogenetic or optogenetic activation of BF cholinergic neurons promotes gamma oscillations, and acetylcholine (Ach) levels in the rodent cortex vary across behavioral states, with the highest levels during attentional tasks and locomotion [65–68]. Importantly, blocking Ach signaling results in missed cues, and selective depletion of Ach in the PFC impairs spatial working memory [69–71]. ACh is thought to enhance the brain’s signal-to-noise ratio; for example, stimulating cholinergic axons in the visual cortex improves the detection of visual stimuli, whereas inhibiting these projections impairs detection [72, 73].

The orexin neurons heavily innervate and excite BF cholinergic neurons, and infusion of orexin-A into the BF increases wakefulness and Ach levels in the cortex [74–76]. Presentation of food is a very salient stimulus in food-restricted rats, resulting in a clear rise in ACh in the PFC, but the destruction of the orexin neurons substantially blunts this rise, highlighting the importance of these neurons [77]. Perhaps in people with narcolepsy, reduced release of ACh from the BF underlies some of the lapses seen on the PVT and SART.

Dopaminergic and noradrenergic neurons

Dopamine (DA) neurons in the VTA and norepinephrine (NE) neurons in the LC promote arousal, attention, memory, and reward, and these two systems likely work in synergy to regulate cognition [78–84]. Dopaminergic and noradrenergic neurons heavily innervate the PFC, but distinguishing their specific contributions is challenging as these two neurotransmitters often show parallel changes [85–87]. Amphetamines primarily block reuptake of DA, but they also block reuptake of NE, and higher levels of both monoamines may contribute to the alerting and attention-enhancing effects of these drugs [88]. Still, they are certainly essential because depleting DA and NE in the PFC produces behavioral consequences comparable to removing the cortex [80].

Motivation is essential for all cognitive tasks, and orexins likely enhance signaling through dopaminergic projections to the cortex. DA neurons in the VTA are activated by orexins primarily through the OX1R, increasing the release of DA in the PFC, nucleus accumbens, and other regions. Substantial research shows that orexin signaling via these mesolimbic pathways promotes drug-seeking and other rewarding behaviors [51]. For instance, systemic administration of an OX1R antagonist reduces the usual surge in DA with cocaine, and rats will not work as hard for cocaine after getting this antagonist [89]. Beyond drugs of abuse, these pathways are probably indispensable for the optimal execution of many goal-oriented tasks. As further examples, mice with acutely reduced orexin signaling will not work as hard for a food reward, and mice lacking orexins run as fast as wild-type mice, but their running bouts are shorter [90, 91]. Hence, activation of dopaminergic pathways by orexins is central to shaping purposeful cognitive and behavioral responses, and low orexin tone in NT1 may underlie some of the feeling that cognitive tasks take more effort.

Orexin neurons also densely innervate noradrenergic neurons of the LC which express only OX1R. The LC is essential for responding to salient stimuli, especially aversive stimuli, and it appears crucial for activating the “bottom-up” ventral attention system [92, 93]. Blocking OX1R signaling in the LC reduces threat learning, whereas photoactivation of orexin terminals in the LC enhances learning [94]. Deleting OX1R from the noradrenergic neurons also reduces theta and gamma EEG activity, especially under conditions that trigger high arousal such as when a mouse is moved to a new cage [95]. These findings suggest that orexin signaling in the LC helps shape attention, especially when attending to salient stimuli, and orexin deficiency in NT1 may lessen LC activity, resulting in impaired attention.

Conclusions and Future Directions

Neuropsychological studies have shown that people with narcolepsy often have difficulty with attention, especially sustained attention, plus possibly some executive dysfunction and memory impairment. Yet, our understanding of the cognitive deficits associated with narcolepsy is still quite limited as most studies have focused on NT1, with small sample sizes, and have tested only a few cognitive domains (e.g. attention and executive function).

Many aspects of cognition would benefit from further investigation in larger groups of individuals with narcolepsy. Furthermore, many unexplored questions remain. How does cognition differ between NT1 and NT2? Does napping produce better improvements than medications? How important is time on task and time of day? To what extent does brain fog overlap with these cognitive impairments? Is cognition worse in children with narcolepsy and does it improve over time?

Importantly, can researchers develop better treatments for narcolepsy-related cognitive impairment? On current wake-promoting medications, patients often report less tendency to actually fall asleep, yet their cognitive symptoms may be only partially improved [19–24, 28]. Many treatments (e.g. modafinil, amphetamines, pitolisant, and solriamfetol) increase DA and other monoamines [96], but might the cognitive problems have a different neurochemical basis that is not fully addressed by these medications?

Preclinical research has shed light on how this impaired cognition may be a consequence of reduced orexin signaling in the PFC, cholinergic, and monoaminergic systems. Orexin receptor agonists hold great promise as highly effective medications for treating sleepiness and cataplexy in NT1 [97], but it remains unknown whether they can address the cognitive symptoms of NT1 or in people with NT2 with normal orexin tone. OX2R is expressed in the PFC, hippocampus, BF, tuberomammillary nucleus, VTA, dorsal raphe, and other areas. However, OX1R is also expressed in many of these same areas, and VTA dopaminergic neurons and LC noradrenergic neurons express only OX1R [57]. If normal attention requires direct activation of the LC, will an OX2R agonist suffice, or might some signaling through the OX1R be necessary? The cognitive response to OX2R agonists could soon be tested in people with NT1 and NT2 as the development of these drugs is well underway, and it would be ideal to compare these to other classes of medications. If some domains of cognition remain problematic, then similar tests with an OX1R agonist would be informative if these medications move into clinical trials.

This future research should provide a fuller understanding of the cognitive problems of narcolepsy, establish better neurobiological models, and provide a rational basis for more effective therapies.

Supplementary material

Supplementary material is available at SLEEP online.

Funding

The authors report no targeted funding.

Disclosure Statement

Financial disclosure: BTH is an employee of Takeda Development Center Americas, Inc., which is developing orexin agonists, and a stockholder of Takeda Pharmaceutical Company, Limited. T.E.S. has consulted for Harmony Biosciences, Jazz Pharmaceuticals, Merck, and Takeda Pharmaceuticals which are developing orexin agonists and has received research grants from the National Institutes of Health, Merck, Jazz, Harmony Biosciences, and Takeda. C.A.C. has no competing financial arrangements or connections that could have influenced the work reported in this paper. Nonfinancial disclosure: The authors declare that they have no known potential conflicts of interest that could have influenced the work reported in this paper.

Data Availability

Data sharing is not applicable to this article as no new data were created or analyzed in this study.