The importance of rapid eye movement sleep and its implications for Alzheimer’s disease

In the study of sleep and neurodegeneration, slow-wave sleep (SWS) has typically garnered more attention due to its connection with Alzheimer’s disease (AD) and other neurodegenerative conditions. Researchers have found that SWS, often referred to as deep sleep, plays a crucial role in memory consolidation and brain health [1, 2]. The relationship between rapid eye movement (REM) sleep and neurodegeneration is less well understood, though some studies suggest that REM sleep is important for cognitive function, emotional regulation, and brain homeostasis, potentially because of synaptic rescaling after SWS [3–7]. However, recent research indicates that dysfunction in REM sleep might have significant implications for cognitive decline and AD [5, 8, 9]. This emerging evidence suggests that disturbances in REM sleep could signal early neurodegenerative changes or contribute to the progression of AD. As researchers explore the impact of REM sleep on neurodegeneration, understanding its role in brain health is becoming increasingly important.

Given this, there’s a growing need to understand the specific functions of REM sleep and the consequences of its dysregulation. Insights into how REM sleep affects AD progression could lead to more comprehensive diagnostic tools and targeted interventions to address sleep-related disruptions in AD. By focusing on REM sleep, researchers can gain a more complete picture of the complex relationship between sleep and neurodegeneration, paving the way for innovative approaches to early detection and treatment.

A recent study by André et al. explored the impact of Apolipoprotein E ɛ4 (APOE4), a significant genetic risk factor for AD, on sleep architecture [10]. They did polysomnography assessments on 109 cognitively unimpaired and 89 amnestic mild cognitively impaired (aMCI) individuals. APOE genotyping on the 198 individuals revealed a cohort of 157 APOE4 non-carriers and 41 APOE4 carriers. Using ANCOVA they did a series of analyses to test the association between APOE4 status and different sleep stages. In the first series of analyses, they covaried for age, sex, cognitive status, and a measure of sleep apnea. In the second series of analyses, they divided the cohorts into sub-groups to account for age, sex, cognitive status, and diagnosis of sleep apnea. The third series of analyses focused on the association of APOE4 status on duration in minutes within each sleep stage as opposed to the percent time in each sleep stage. The fourth analysis focused on the effect of sleep apnea specifically during REM sleep as well as repeating analyses after removing individuals with moderate-to-severe sleep apnea. The study revealed that APOE4 carriers experienced reduced REM sleep (duration in both minutes and percent time), while the other sleep stages were not associated. The authors convincingly showed that the APOE4-associated reduction in REM sleep was not driven by the effect of sleep apnea which is known to disrupt REM sleep.

There were several important strengths of this study. They used gold-standard polysomnography to assess sleep in a large cohort of individuals. The cohorts in this study were well classified, using neuropsychological assessments to identify individuals as cognitively healthy, cognitively impaired in the memory domain, or cognitively impaired in non-memory domains. Furthermore, the authors were very thoughtful in their approach to separate the contaminating influence of sleep apnea on REM disruption from the APOE4-associated diminished REM sleep duration. The limitations were standard ones that affect this kind of research, with a limited sample size (particularly for APOE4 carriers) and a cross-sectional approach. As the authors highlighted, it would be relevant to examine the sleep association of one vs two APOE4 alleles, though the sample studied did not have a sufficient sample size of APOE4/4 carriers. It would also be important to establish if the REM/APOE4 association occurred throughout life or is specific to older adulthood.

Though there are mixed findings on the exact role that REM sleep has in memory consolidation and emotional regulation, it is generally thought to have a pertinent one [4, 5]. The observed reduction of REM sleep in APOE4 carriers is concerning [10], suggesting that it might signal early neurodegenerative processes in regions responsible for REM sleep generation and maintenance. This observation underscores the potential for REM sleep dysfunction to serve as an early indicator of Alzheimer’s progression. As André et al. highlighted, they are not the first to identify a link between diminished REM sleep duration in APOE4 carriers [11] though subsequent studies have produced mixed findings (e.g. Drogos et al., 2016) [12]. Other work done by Baril et al., in 2020 also linked REM sleep to neural health, where they described an association between lower time spent in REM sleep and diminished subcortical gray matter in APOE4 carriers that was not identified in non-carriers [13]. Our own group has previously described an association between REM sleep and white matter integrity in cognitively healthy older adults when covarying for demographics as well as APOE status and sleep apnea [14]. This study found that the percentage of time spent in REM sleep was positively associated with white matter integrity, reinforcing the role of REM sleep in maintaining brain health. To provide a cognitively relevant framework, in APOE4 asymptomatic individuals, memory-guided attention, dependent on network communication between memory and attention areas, was impaired [15]. Though memory itself was not impaired in these APOE4 carriers, a task that taps into multiple centers with reliance on the use of white matter was impaired.

Moreover, in a recent study, we found that typical Alzheimer’s disease (characterized primarily by memory impairment) exhibits distinct patterns of sleep disturbances compared to the atypical variants that manifest with language or visuospatial deficits [16]. The research, involving 48 participants across different AD phenotypes, showed that typical AD is associated with greater impairment in SWS, while atypical AD displays more deficits in REM sleep, suggesting that sleep dysfunction might contribute to specific patterns of cognitive dysfunction in AD. On the other hand, REM sleep dysfunction has been associated with specific brain changes in AD. For instance, associations between REM sleep dysfunction and brain changes have been further emphasized by Oh et al., who demonstrated a correlation between REM sleep dysfunction and subcortical structures like the locus coeruleus, one of the earliest brain areas affected by tau pathology in AD [17]. Interestingly, there appears to be an interaction between REM sleep behavior disorder (RBD) and APOE4, which is associated with increased sundowning in AD [18]. It is unclear if APOE4 is associated with altered REM sleep network dynamics or if specific nuclei may be affected. This connection between REM sleep and subcortical neurodegeneration further underscores the relevance of REM sleep in AD.

Together, these studies indicate that REM sleep is crucial in the progression of AD and overall brain structure integrity. Further research into REM sleep disruptions and their contribution to neurodegeneration could lead to new approaches for early detection, diagnosis, and targeted treatment strategies in AD. Finally, delving into the effects of REM sleep impairment in neurodegenerative diseases could help us unlock the pivotal role REM sleep plays in brain health, revealing its significance far beyond dreaming.

Disclosure statement

Financial disclosure: NF is supported by Alzheimer’s Association grant (AACSF-21-723056) and the Carlos III Health Institute with European funds from the Recovery, Transformation and Resilience Plan, with file code JR22/00014, by virtue of the Resolution of the Carlos III Health Institute Management, O.A., M.P. of December 7, 2022, by which the Juan Rodes Contracts are granted, and “Funded by the European Union – NextGenerationEU. Nonfinancial disclosure: None.

Data availability

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