Short sleep duration and dementia: a narrative review

Abstract

Short sleep duration has been linked to the development of neurocognitive disorders. Still, current evidence for this relationship is conflicting. In this review, we summarize evidence regarding the relationship between short sleep duration and neurocognitive disorders, which shows that short sleep duration increases the risk of incident major neurocognitive disorders beginning as early as midlife. The pathological brain changes attributed to poor sleep may be related to changes in brain microstructure and accumulation of debris in the brain. More evidence is needed to fully understand the relationship between sleep duration and cognitive decline and the molecular changes that link the two. Measures of sleep quality such as sleep duration represent a potentially modifiable risk factor for the prevention of cognitive decline and neurocognitive disorders.

Several studies suggest a link between sleep duration and the development of neurocognitive disorders (NCD). Thus, improvements in sleep quality may mitigate the onset of NCD, particularly in aging populations. A V- or U-shaped association between sleep and incident NCD, where both long and short sleep duration contribute to an increased risk for cognitive decline, has been found in study populations from North America,¹ China,² Japan,³ Finland,⁴ and Sweden,⁵ as well as Ghana, India, Russia, South Africa, and Mexico.⁶ Other studies have expanded upon the effect of short sleep duration and the onset of major NCD and are likewise consolidated in this review.

METHODS

We conducted a thorough search through PubMed and Scopus databases using the keywords “sleep duration” and “dementia.” We narrowed down the search results based on full-text availability and English translation and reviewed title and abstract for articles published in the last 30 years (January 1990 to May 2021), selecting papers that directly addressed a potential link between sleep duration and the development of major NCD and cognitive decline, not including Parkinson’s disease, with emphasis on short sleep duration as a risk factor. Sleep-related exposures other than duration (e.g., insomnia, sleep-disordered breathing, etc.) were excluded from this review.

RESULTS

Supporting and contrasting evidence for the association between sleep duration and cognitive impairment is shown in Table 1.^1–12

Table 1.

Summary of major findings on the relationship between sleep duration and neurocognitive disorders

Study	Study type	Sleep measures	Outcomes	Main findings
Xu et al, 20201	Meta-analysis of 17 longitudinal prospective cohort studies	Nocturnal sleep duration*	Cognitive impairment, major NCD, AD, VD	V-shaped association between sleep duration and cognitive decline, major NCD, and AD; significantly increased risk associated with habitual nighttime sleep <4 and >10 hours in a dose-response manner
Wu et al, 20212	Prospective cohort study, N = 16,948	Total sleep duration*	Cognitive impairment	U-shaped association between sleep duration and MMSE score; change in sleep duration (increased/decreased) associated with increased odds of cognitive impairment
Ohara et al, 20183	Prospective cohort study, N = 1517	Total sleep duration*	Major NCD, VD, AD	Increased risk for major NCD, AD, and VD associated with sleep duration <5, 8–9.9, and >10 hours
Virta et al, 20134	Prospective cohort study, N = 2336	Nocturnal and total sleep duration*	Cognitive impairment, AD	Habitual short (<7 hours) and long (>8 hours) sleep duration starting in midlife associated with cognitive impairment; long sleep associated with increased risk for AD
Bokenberger et al, 20175	Register-based cohort study, N = 11,247	TIB, rise time, and bedtime*	Major NCD	U-shaped association for TIB (<6, >9 hours) and incident major NCD; short TIB is a risk factor for major NCD
Gildner et al, 20146	Stratified multistage cluster, N = 34,203	Nocturnal sleep duration*	Cognitive impairment	Short (<6 hours) and long (>9 hours) sleep associated with lower cognitive performance
Robbins et al, 20217	Longitudinal cohort study, N = 2812	Nocturnal sleep duration*	Major NCD	Short (<5 hours) sleep associated with a twofold increased risk for major NCD
Chen et al, 20168	Prospective cohort study, N = 7444	Nocturnal sleep duration*	Minor and major NCD	Short (<6 hours) and long (>8 hours) sleep associated with increased risk for NCD; short sleep resulted in 30% increased risk after adjustment for confounders
Sabia et al, 20219	Longitudinal prospective cohort study, N = 8321	Nocturnal sleep duration* accelerometer	Major NCD	Short sleep (<6 hours) associated with a greater risk for incident major NCD in adults aged 50–65 preceding average onset of major NCD (75 years)
Ferrie et al, 201110	Longitudinal prospective cohort study, N = 10,308	Nocturnal sleep duration* accelerometer	Cognitive impairment	V-shaped association between sleep duration and cognitive impairment; sleep-related changes in cognitive decline equivalent to 4–7 years of aging
Lu et al, 201811	Population-based cohort study, N = 7422	Total sleep duration*	Major NCD	No association between short sleep duration and incident major NCD
Lysen et al, 201812	Prospective population-based study, N = 4835	Nocturnal sleep duration*	Major NCD	No association between short sleep duration and incident major NCD

Conclusions about short and long sleep are in comparison to recommended optimal sleep (7 hours). AD indicates Alzheimer’s disease; MMSE, Mini-Mental State Exam; NCD, neurocognitive disorders; TIB, time in bed; VD, vascular dementia.

^*Self-reported.

DISCUSSION

In this review, we summarize the evidence linking habitual short or decreased sleep duration with risk of developing minor and major NCD. Much of the literature agrees that abnormal sleep duration is associated with higher risk. Thus, sleep duration represents a potentially modifiable risk factor for the prevention of cognitive disorders in later life.

Aging appears to contribute to poorer sleep patterns that precede declining cognitive performance and the development of major NCD and related disorders. Several mechanisms have been described for physiological brain changes that may accompany or precede the onset of major NCD in individuals with short sleep duration. Sleep <6 hours is associated with changes in white matter quality that are detectable earlier than cognitive presentations of dementia and that resemble the neurodegenerative changes seen in Alzheimer’s disease.¹³ These findings were challenged in a 28-year prospective study on changes in brain microstructure and cognition, where there were no differences between groups of varying sleep duration; however, this was attributed to the limited number of study participants with “extreme” sleep duration (<5 hours).¹⁴ An alternative explanation proposed by Zitser et al is that both sleep duration and quality are important for cognitive health; they found healthier looking brains in groups with higher quality short sleep.¹⁴

Sleep may also be important for the clearance of debris from the brain. In mice and human studies, interstitial levels of β-amyloid fluctuated diurnally and were inversely correlated with time spent asleep.¹⁵ Spira et al also found short sleep duration and poor sleep quality were associated with greater levels of β-amyloid, even in the absence of NCD.¹⁶ After only a single night of sleep deprivation, β-amyloid levels were significantly elevated in the hippocampus and thalamus in humans—brain regions relevant in Alzheimer’s disease pathology.¹⁷ One potential mechanism for this association involves the capacity for cerebrospinal fluid exchange, which is enhanced during sleep, allowing for the clearance of toxic metabolites.¹⁸ Aggregation of other metabolites (e.g., tau, alpha-synuclein) is associated with sleep deprivation and the development of other NCD.¹⁹

Much of the literature still has not reached an agreement on the most accurate approach to studying sleep duration. Major NCD and related cognitive disorders are slow-progressing diseases that take many years to develop before the manifestation of symptoms. Thus, much of the work cited here may be limited by short follow-up time, and older participants may have had subclinical major NCD at enrollment. Moreover, self-reported sleep duration is a highly subjective measure of true sleep duration and may contribute to skewed results. Subjective sleep data is best supplemented by objective measures of sleep duration (e.g., actigraphy, accelerometer).²⁰ Disrupted sleep patterns are commonly found in individuals with major NCD, and it can be difficult to determine which of the two plays the causal role. Studies with earlier onset (younger participants) and longer duration of follow-up are less likely to produce results that are skewed by reverse causation. Excluded from this review were many reports that included sleep disorders other than sleep duration (e.g., sleep apnea and/or disordered breathing, difficulty falling asleep, napping), which may represent more specific risk factors for major NCD and may be concomitant with shorter sleep duration. Few of the sources discussed here addressed psychiatric comorbidities that may impact both sleep and cognitive function, yet these should be ruled out as confounders when possible.

Sleep duration is linked not only to cognitive decline but to many other disorders that may contribute to the development of major NCD. Thus, improving sleep habits may impact the onset of major NCD directly or indirectly by improving related health conditions. Moreover, primary care screening for sleep disorders and early intervention may mitigate the risk of developing major NCD later in life. Polysomnography should be considered more frequently in individuals presenting with sleep issues. Pharmacological interventions (e.g., trazodone) and nonpharmacological interventions (cognitive behavioral therapy) may be useful for early intervention to improve sleep conditions and lower the risk of cognitive decline.^21,22 Clinical trials assessing the efficacy of early interventions is warranted. Despite the role of sleep in the development of major NCD and related disease, it remains possible that adverse trends in metabolic and cardiovascular health—known risk factors for cognitive decline—will truncate the effects of improved sleep health in groups with comorbidities.²³

In conclusion, sleep duration is implicated in the development of major NCD and represents a potentially modifiable risk factor. Thus, optimal sleep duration should be encouraged to mitigate the risk of cognitive decline later in life. While many hypotheses exist for the mechanism linking short sleep to major NCD, further research is warranted to build evidence for these relationships.