People with moderate to severe Alzheimer’s disease (AD) may be kept awake by a neurotransmitter that goes bump in the night. According to a study published October 13 in JAMA Neurology, cerebrospinal fluid levels of orexin, a peptide neurotransmitter that staves off shut-eye, are elevated in patients with moderate to severe AD. This correlated with bouts of wakefulness in the wee hours, and also with CSF tau, a marker of neurodegeneration. Led by Fabio Placidi at the University of Rome Tor Vergata, Italy, the study suggests that an abundance of orexin may play a part in AD-associated sleep disturbances, which some researchers believe may further promote progression of the disease.

“This study significantly adds to our understanding about the relationships among sleep, orexin, and AD,” wrote Brendan Lucey of Washington University in St. Louis in an email to Alzforum. “It provides additional evidence that sleep is disturbed in moderate to severe AD and that the orexinergic system is involved.”

Sleep disturbances often precede and accompany AD, and may even hasten its onset, though it is not clear how. Lucey’s colleagues at WashU, including Randall Bateman and David Holtzman, have reported that Aβ levels in the CSF drop during a normal night's slumber, whereas in people deprived of sleep it holds steady (Bateman et al., 2007, and Jun 2014 news story on Ooms et al., 2014). In mice, lack of sleep accelerates Aβ deposition, and orexin reportedly boosts Aβ levels as well (see Sep 2009 news story on Kang et al., 2009). 

Lucey and colleagues have proposed that this potentially toxic relationship between Aβ deposition and sleep deprivation may accelerate AD pathology in humans as well, and that the neurodegeneration that ensues may further damage parts of the brain essential for a good night’s sleep (see Lucey and Bateman, 2014, and Ju et al., 2014). Overly active orexinergic neurons in the hypothalamus could be one way in which neurodegeneration keeps people awake. Through their far-reaching axon projections, this small population of neurons delivers this neuropeptide (also known as hypocretin) to multiple brain regions, where it promotes wakefulness. Narcoleptics have abnormally low orexin levels, though this does not protect from AD, according to one small study (see Scammell et al., 2012).

Other studies found no differences in CSF orexin levels between people with AD and healthy controls, whereas one study found that people with mild cognitive impairment or AD had elevated orexin levels compared to people with other forms of dementia, such as frontotemporal lobar degeneration and dementia with Lewy bodies (see Dauvilliers et al., 2014). 

To help clarify orexin’s role in AD, first author Claudio Liguori and colleagues measured CSF levels of the neuropeptide in people in varying stages of the disease, and also factored in measurements of sleep quality and cognition. The study included 21 people with mild AD, 27 with moderate to severe AD, and 29 healthy controls. The researchers asked volunteers and their caregivers to record their nightly sleep experiences in a diary for one week, and then over two nights placed polysomnography electrodes on them to measure how much and how well they slept. After the second night, the researchers collected CSF samples via lumbar puncture. As expected, people with AD had higher levels of tau and p-tau, and lower levels of Aβ42, than did controls. People with AD also slept poorly—they took longer to enter REM sleep and woke up more often after falling asleep than did controls. Those with moderate to severe disease slept worst of all, according to both polysomnography and sleep diaries. CSF orexin levels did not differ from controls when considering the AD group as a whole. However, people with moderate to severe AD had higher orexin, suggesting that it ramps up as the disease progresses.

The researchers next looked for correlations between AD biomarkers, orexin, sleep quality, and cognition. Levels of CSF total tau rose with orexin levels in both AD groups, as did p-tau in the moderate to severe AD group. No link was found between Aβ42 and orexin in any group; however, waning CSF Aβ42 levels and higher orexin levels each correlated with reduced sleep quality. Scores on the mini mental state exam (MMSE) did not track with orexin or any AD biomarker, but in both AD groups those with the worst sleep patterns had the poorest cognitive scores.

How do these findings mesh with those of previous studies? Three prior studies found no relationship between CSF orexin and AD, but they did not stratify groups by disease severity and had fewer participants (see Wennstrom et al., 2012Slats et al., 2012; and Schmidt et al., 2013). They may have missed the emergence of high orexin levels later in the disease, Ligouri said. Indeed, when he re-analyzed results from one of the studies (Schmidt et al., 2013), he found a trend toward higher orexin levels in patients with moderate to severe cognitive impairment. A fourth study reported lower levels of orexin and orexinergic neurons in postmortem brains from AD patients. Liguori and colleagues speculated that those patients had more advanced stages of the disease (Fronczek et al., 2013). 

While Liguori and colleagues observed a link between CSF tau and orexin levels, they did not see a connection between CSF Aβ and orexin. This conflicts with Slats’ 2012 study on a small number of AD patients. Liguori concluded that in his cohort, CSF Aβ levels had already bottomed out; indeed, he found no difference in Aβ levels among patients with mild and moderate to severe AD.

“The results clearly demonstrate and support a role for dysregulation of the orexinergic system in AD,” wrote Henrietta Nielsen of the Mayo Clinic in Jacksonville, Florida, in an email to Alzforum. Nielsen, who led one of the studies that found no connection between orexin levels and AD, wrote that to understand the underlying molecular pathways, studies will have to combine antemortem CSF and postmortem brain tissue analysis of patients who progressed from mild cognitive impairment to AD.

Are rising orexin levels a consequence of AD, or an early part? Liguori believes that orexin’s rise may help drive the disease. “The increased orexinergic tone seems to cause poor sleep quality, and it is well known that sleep dysregulation alters cognition,” he wrote in an email to Alzforum, concluding, “In AD patients, impaired sleep may accelerate cognitive decline.” The authors hypothesized that a breakdown in cholinergic signaling, which normally dials down orexin output, may open the floodgates of orexin release in the context of AD.

In contrast, Lucey believes that abnormal orexin levels are likely a consequence of AD. “Given the lack of correlation between CSF orexin concentration and mild AD, I think the study suggests sleep disturbance is more likely to be a result of AD rather a key part of AD pathogenesis,” he wrote. He added, however, that a larger group size could have teased out the beginnings of orexin’s rise in mild cases.

Either way, Liguori and colleagues propose that blocking the neurotransmitter with receptor antagonists could at least help AD patients sleep better at night. In an accompanying editorial, Luigi Ferini-Strambi of the Vita-Salute San Raffaele University in Milan proposed to go even further. “It could be hypothesized that the use of orexin receptor antagonists as potential drugs targets the downregulation of the orexinergic system not only for the management of sleep disturbances in AD, but also for a slower progression of the neurodegenerative process,” Ferini-Strambi  wrote.  In August, Merck’s Belsomra became the first orexin receptor antagonist approved by the FDA for the treatment of insomnia.—Jessica Shugart


  1. This study significantly adds to our understanding about the relationship between sleep, orexin, and AD. The authors measure sleep more robustly than in other published studies with home sleep patterns reported via sleep diaries and an acclimation polysomnogram preceding the study polysomnogram. This is important because if there was only one night of polysomnography we could not say how representative it was of the participants' sleep at home. Further, the lumbar punctures were performed within a narrow time period.

    The main findings of the study are: 1) CSF orexin concentration positively correlated with t-tau and p-tau; 2) CSF orexin levels correlated with several sleep parameters (wakefulness after sleep onset, sleep efficiency, percent of REM sleep, and sleep latency). The study provides additional evidence that sleep is disturbed in moderate-severe AD and that the orexinergic system is involved.

    Given the lack of correlation between CSF orexin concentration and mild AD, I think the study suggests sleep disturbance is more likely to be a result of AD rather a key part of AD pathogenesis. Sleep disturbance in AD may be another measure of brain dysfunction, such as cognitive impairment. A 2013 study by my colleagues Drs. Ju and Holtzman at Washington University showed that cognitively normal adults with amyloid deposition had reduced sleep efficiency compared to cognitively normal adults without amyloid deposition; this study had a much large number of participants and showed a small but statistically significant reduction in sleep efficiency (Ju et al., 2014). It may be that larger sample sizes are needed to define differences in CSF orexin concentrations between the control and mild AD group.


    . Sleep and Alzheimer disease pathology--a bidirectional relationship. Nat Rev Neurol. 2014 Feb;10(2):115-9. Epub 2013 Dec 24 PubMed.

  2. Liguori and colleagues report increased orexin in CSF from patients with mild to moderate Alzheimer’s disease (MMSE score <21). In parallel, these patients also had impaired nocturnal sleep compared to controls and patients with mild AD (MMSE score ≥ 21). Interestingly, the authors found no significant difference between AD patients with mild versus those with moderate to severe AD. However, the numbers reveal an increase in CSF orexin by approximately 12 percent in the moderate to severe versus mild AD patients, hence if the authors had included more patients (increased the power) this difference may have been significant.

    No gender-associated differences were present in the investigated cohort. In regard to sleep efficiency and cognitive performance, the authors, not too surprisingly, reported a positive association between sleep efficiency and MMSE scores in AD patients, and their findings of negative correlations between sleep efficiency as well as REM sleep percentage and CSF orexin levels validate the biological links between orexin levels and sleep patterns. Importantly, the authors also investigated potential links between CSF orexin and Ab42 levels and found no evidence thereof. Hence, this study could not confirm the earlier described link between orexin and Aβ in mice (Kang et al., 2009). Liguori et al., however, speculate that the lack of correlation may be due to the fact that Aβ42 levels in their AD patients may already have reached a plateau, as they found no difference in Aβ levels between AD patients with different disease severities. Hence, the lack of association between CSF Aβ and orexin levels doesn’t out-rule a biological link because the authors did, in fact, report that decreased Aβ levels were associated with sleep deterioration in a sample of their AD cohort with severity ranging from mild to moderate disease.

    The study contains a detailed and well-performed polysomnographic assessment of various sleep variables in AD patients. Their findings, overall, add to the growing body of evidence supporting the notion of a dysregulated orexinergic system in AD patients, who frequently experience disturbed sleep patterns. The results are in line with previous reports, although results from different studies are at times conflicting. For instance, whereas Liguori et al. found no gender-associated differences in CSF orexin levels, we and others have reported increased levels in females versus males (Wennström et al., 2012; Schmidt et al., 2013). Gender-associated differences in sleep patterns were previously reported in a large study by Silva and colleagues (Silva et al., 2008). In our own study, we looked at AD patients with mild dementia (MMSE score ≥ 21), and similar to Liguori et al., we could not detect altered orexin levels when compared to controls in groups of mixed gender. Female AD patients, however, did present significantly higher CSF orexin levels compared to controls (Wennström et al., 2012). Hence CSF orexin levels may not be sensitive to pathological changes until these become severe and group comparisons may have to be strictly stratified for confounding factors like gender. The interpretation of the results from human studies are complicated, as basic descriptive studies are lacking, for instance in regard to potential gender differences in the number of orexin-producing neurons in the hypothalamus, as well as how decreased CSF orexin levels relate to neurodegeneration. The results presented by Liguori et al. clearly demonstrate and support a role for a dysregulation of the orexinergic system in AD, but in order to further improve our understanding of the underlying molecular pathways, studies combining postmortem brain tissue and antemortem CSF analysis of patients progressing from mild cognitive impairment to AD are crucial. Also, since individuals carrying the APOE4 allele are known to accumulate Aβ pathology even in the absence of cognitive symptoms (Morris et al., 2010), it would be interesting to include APOE4-positive narcolepsy patients, in which a specific loss of orexin-producing cells is paralleled by decreased CSF orexin levels, in future studies of potential associations between orexin, wakefulness, Aβ levels, and plaque pathology.


    . Amyloid-beta dynamics are regulated by orexin and the sleep-wake cycle. Science. 2009 Nov 13;326(5955):1005-7. PubMed.

    . Altered CSF Orexin and α-Synuclein Levels in Dementia Patients. J Alzheimers Dis. 2012 Jan 1;29(1):125-32. PubMed.

    . Cerebrospinal fluid melanin-concentrating hormone (MCH) and hypocretin-1 (HCRT-1, orexin-A) in Alzheimer's disease. PLoS One. 2013;8(5):e63136. Print 2013 PubMed.

    . Gender and age differences in polysomnography findings and sleep complaints of patients referred to a sleep laboratory. Braz J Med Biol Res. 2008 Dec;41(12):1067-75. PubMed.

    . Altered CSF Orexin and α-Synuclein Levels in Dementia Patients. J Alzheimers Dis. 2012 Jan 1;29(1):125-32. PubMed.

    . APOE predicts amyloid-beta but not tau Alzheimer pathology in cognitively normal aging. Ann Neurol. 2010 Jan;67(1):122-31. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. While You Were Sleeping—Synapses Forged, Amyloid Purged
  2. Sleep Deprivation Taxes Neurons, Racks Up Brain Aβ?

Paper Citations

  1. . Fluctuations of CSF amyloid-beta levels: implications for a diagnostic and therapeutic biomarker. Neurology. 2007 Feb 27;68(9):666-9. PubMed.
  2. . Effect of 1 night of total sleep deprivation on cerebrospinal fluid β-amyloid 42 in healthy middle-aged men: a randomized clinical trial. JAMA Neurol. 2014 Aug;71(8):971-7. PubMed.
  3. . Amyloid-beta dynamics are regulated by orexin and the sleep-wake cycle. Science. 2009 Nov 13;326(5955):1005-7. PubMed.
  4. . Amyloid-β diurnal pattern: possible role of sleep in Alzheimer's disease pathogenesis. Neurobiol Aging. 2014 Sep;35 Suppl 2:S29-34. Epub 2014 May 15 PubMed.
  5. . Sleep and Alzheimer disease pathology--a bidirectional relationship. Nat Rev Neurol. 2014 Feb;10(2):115-9. Epub 2013 Dec 24 PubMed.
  6. . Coexistence of narcolepsy and Alzheimer's disease. Neurobiol Aging. 2011 Jan 21; PubMed.
  7. . Hypocretin and brain β-amyloid peptide interactions in cognitive disorders and narcolepsy. Front Aging Neurosci. 2014;6:119. Epub 2014 Jun 11 PubMed.
  8. . Altered CSF Orexin and α-Synuclein Levels in Dementia Patients. J Alzheimers Dis. 2012 Jan 1;29(1):125-32. PubMed.
  9. . Association between Hypocretin-1 and Amyloid-beta 42 Cerebrospinal Fluid Levels in Alzheimer�s Disease and Healthy Controls. Curr Alzheimer Res. 2012 Jun 26; PubMed.
  10. . Cerebrospinal fluid melanin-concentrating hormone (MCH) and hypocretin-1 (HCRT-1, orexin-A) in Alzheimer's disease. PLoS One. 2013;8(5):e63136. Print 2013 PubMed.
  11. . Hypocretin (orexin) loss in Alzheimer's disease. Neurobiol Aging. 2011 May 3; PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Orexinergic system dysregulation, sleep impairment, and cognitive decline in Alzheimer disease. JAMA Neurol. 2014 Dec;71(12):1498-505. PubMed.