Sleep disturbances and dementia



Sleep is a complex behavioural state, the ultimate functions of which remain poorly understood. It becomes more fragmented as we age, with more night-time awakenings and greater tendency for daytime sleep. The magnitude of disordered sleep among individuals affected by dementia has been clearly demonstrated, and disturbed sleep is a major clinical problem in dementia. Comorbid insomnia and other sleep disturbances are common in patients with neurodegenerative disorders, such Alzheimer's disease and other dementing disorders. How and when sleep problems manifest themselves can depend on the type of dementia involved as well as the stage of the dementia. However, differences in sleep pattern presentation show more variation during the initial stages of dementias than they do during the later stages. Effective, pragmatic interventions are largely anecdotal and untested.


Sleep is a complex phenomenon that is rooted in neurologic function. It is an active process generated and modulated by a complex set of neural systems located mainly in the hypothalamus, brainstem, and thalamus. Sleep serves a restorative function in the brain and has a critical role in cognitive functions. Fortunate are those who rise out of bed to greet the morning light well rested with the energy and enthusiasm to drive a productive day.[1] Disturbed sleep can mean different things to different people. Nearly half of older adults report difficulty initiating and maintaining sleep. Sleep decreases physiologically in quantity and quality with age,[2, 3] and insomnia and excessive daytime sleepiness are frequently reported in the elderly.[4] Sleep becomes more fragmented as we age, with more night-time awakenings and greater tendency for daytime sleep.[5] Comorbid insomnia and other sleep disturbances are common in patients with neurodegenerative disorders, such Alzheimer's disease (AD) and other dementing disorders. Furthermore, studies of incident dementia suggest that sleep problems increase the risk of dementia.[6-8]


Not all dementia patients develop sleep problems. According to Guarnieri et al.,[9] individuals suffering from dementia with Lewy bodies (DLB) and Parkinson's disease (PD) with dementia have the highest frequency of occurrence of any sleep disturbances, with 90% of patients affected. The same authors observed that insomnia frequency was identical in AD and frontotemporal dementia (FTD) patients but was about 2.5 and 1.5 times more frequent in vascular dementia (VaD) and DLB/PD with dementia patients, respectively. Nocturnal and daytime sleep disturbances are common in persons with AD, affecting up to 44% of patients in clinic and community-based samples.[10, 11] Using logistic regression analyses, McCurry et al. observed that the factors most strongly associated with night awakenings among patients with AD were male gender, greater memory problems, and decreased functional status.[10]

Disorder Outcomes

Sleep disturbances negatively affect patients' quality of life and functional abilities,[12] and they are associated with a greater risk of psychiatric symptoms.[13] Poor sleep among persons with AD may lead to increased daytime irritability and decreased attention, motivation, and cognitive performance.[14, 15] Furthermore, poor sleep results in an increased risk of significant morbidities and even mortality in demented patients.[16] In fact, sleep problems, or more specifically, sleep debt, may represent one of the biological pathways associated with sympathovagal imbalance and elevated heart rate, leading to early mortality in middle-aged subjects.[17] Also, the rate of falls in older persons with disordered sleep has been shown higher than in healthy older persons.[18]

In addition to being prominent in neurodegenerative diseases, sleep disturbances could exacerbate a fundamental process leading to neurodegeneration: researchers found a relationship between sleep deprivation for just 3 weeks and accelerated development of amyloid plaques (aggregation of extracellular amyloid-β, which is thought to play a major part in the pathogenesis of AD) in the brains of lab mice.[19]

When patients with long-term neurological illnesses have disrupted sleep, this becomes a problem, not only for the patient, but also for their carers. Nocturnal sleep disruption is reported as a major source of caregiver physical and psychological burden.[20] In some cases, caretakers may even develop clinical depression as a result.[21] Furthermore, sleep disturbances are often the catalyst and the primary reason for institutional care.[22-24] Institutionalization has been shown to exacerbate sleep–wake disruption,[25] and to increase the speed of deterioration.[26] Even in nursing homes, circadian rhythm disturbances in AD patients can be disruptive for the staff. In addition, caregivers experiencing emotional or mental strain in conjunction with their caregiving role have been reported to have an increased risk of mortality when compared with non-caregiving controls.[27]


The origin of sleep disorders in dementia is usually multifactorial, resulting from pathophysiological changes associated with the disease itself and environmental factors.[28] For example, in a nursing home, noise and light exposure occur intermittently throughout the night and contribute to sleep disruption.[29] Other factors involve medical or psychiatric morbidity. Medications used to lessen the negative behavioural symptoms of dementia and to slow disease progression are often associated with side effects that negatively affect sleep and wakefulness. It is also known that longer periods in bed are associated with more fragmented sleep;[30] therefore, the chronic bed rest that often characterizes the routine care of AD patients, particularly patients at more advanced stages, may in itself contribute to insomnia.[31] It is highly plausible that the diffuse brain damage that characterizes dementia might extend over the cognitive brain areas to involve the neural networks that control sleep function (i.e. the anterior hypothalamus, reticular activating system, suprachiasmatic nucleus, and pineal gland).[32] The hypothalamic suprachiasmatic nucleus (SCN) is the central biological clock of the brain; it generates and synchronizes the overt biochemical, physiological, and behavioural rhythms throughout our body.[33] It is thought to induce a single circadian oscillation in mammals.[34] The SCN has input pathways that link it to the retina, limbic forebrain, other hypothalamic nuclei, the raphe nuclei, and reticular formation, and it also responds to various hormones. Light entering the retina travels along the optic nerves to the SCN, which triggers the pineal gland to stop producing the neurohormone melatonin, an essential component in sleep; its production is highest during the night, when light stimuli are minimal or absent. Disruptions anywhere along this pathway can cause disruptions in the circadian rhythm and, ultimately, sleep disturbances.[35] Dementia patients have been noted to have abnormalities in their rhythms of melatonin secretion.[36] This dysfunction has been noted not only in patients with clinical diagnosis of AD,[37] but it has also been confirmed after post-mortem analysis. The observed functional disconnection between the SCN and the pineal gland from the earliest AD stage onwards seems to account for the pineal clock gene and melatonin changes, and it underlies circadian rhythm disturbances in AD.[38] Though not clear, genetic risk factors, such as in AD patients who are negative for the APOε4 allele, have also been implicated in the development of sleep problems.[39] There have been few studies of sleep per se in relation to apolipoprotein E status, and the findings are somewhat fragmentary. According to Yesavage et al.,[40] apolipoprotein E status was associated with the progression of sleep–wake disturbances in AD. They observed a greater deterioration on sleep parameters in patients negative for the ε4 allele.

Sleep Disturbances and Dementia Subtypes

Sleep disorders seem to be different in each form of dementia, and this may be a consequence of how the brain's varying pathological involvement characterizes each type of dementia.[41-43] However, few studies have compared the presence of disturbed sleep between neurodegenerative conditions. How and when sleep problems manifest themselves can depend on the type of dementia involved as well as the stage of dementia; differences in sleep pattern presentation show more variation during the initial stages of dementia than they do during the later stages. The pathology of AD (amyloid-β accumulation in the brain) emerges prior to any symptoms, with the first identifiable changes occurring ∼10–15 years before cognitive symptoms. Changes in sleep seem to precede the onset of cognitive symptoms in patients with AD, and the further decline of sleep quality and/or circadian function parallels both cognitive dysfunction and the progression of AD pathology. This is consistent with the finding that brain regions involved in sleep and circadian control are affected early in the pathogenesis of the condition.[44] In a population-based sample of AD, the most common sleep-related behaviour problems reported by caregivers were sleeping more than usual (40%) and awakening early (31%), whereas being awakened at night (24%) was the most distressing problem for caregivers.[10] Nocturnal sleep disturbance in AD patients is often accompanied by increased daytime napping,[45] frequently in direct association with the extent of dementia.[46, 47]

DLB is a common form of dementia in old age, sharing clinical and pathological features with both AD and PD. Sleep profiles in DLB differ from those in AD. Both groups suffer from sleep problems, but DLB patients seem to suffer from more overall sleep disturbance. Whether synucleinopathies or amyloidopathies are the neurobiological substrates for the preferential association of sleep disturbance remains unknown, but they are thought to involve the primacy of brainstem degeneration in parkinsonian-like conditions.[48, 49] DLB patients have a greater tendency to fall asleep at inappropriate times during the day and to also have more night-time sleep disturbances.[50] Ferman et al. found characteristics to significantly differentiate AD from DLB.[51] Among the composite features, they included daytime drowsiness and lethargy and daytime sleep of 2 h or more.

Rapid eye movement sleep behaviour disorder (RBD) occurs frequently in patients with α-synuclein pathology, including DLB, PD, and multiple system atrophy.[52] RBD is characterized by repeated episodes of arousal during sleep associated with vocalization and/or complex motor behaviours, which may be sufficient to result in injury to the individual or bed partner. Affected patients have excessive motor activity such as punching, kicking, or crying out in association with dream content.[53] Most patients view their dreams as nightmares, and the dream content often involves insects, animals, or people chasing or attacking them, their relatives, or their friends.[52]

In idiopathic RBD, the risk of developing neurodegenerative disease is substantial. The interval between RBD onset and disease onset averages 10–15 years, suggesting a promisingly large window for intervention into preclinical disease stages.[54] Among the types of sleep disorders that patients with PD can experience (insomnia, parasomnias, sleep fragmentation, increased daytime drowsiness),[55] there is particular interest in impairments associated with the rapid eye movement (REM) phase. Dementia is seen in 40–70% of PD patients, and it typically occurs 10 years or more after the initial motor signs. Sleep disorders during the REM phase may be combined with cognitive impairments or precede their development. Marion et al. reported a study of 65 patients with PD complicated and not complicated by dementia and found that 77% of patients with dementia had sleep impairments during the REM phase, while behavioural impairments were present in only 27% of patients without dementia.[56] Although RBD secondary to a neurodegenerative disease is commonly associated with synucleopathies (PD, DLB, and multiple system atrophy), it has also been reported in patients with AD.[53]

In 2012, RBD was described in a patient with FTD for the first time.[57] Anderson et al. first demonstrated significant sleep–wake disturbance in non-institutionalized FTD patients, which differed from that seen with AD. FTD subjects showed increased nocturnal activity and decreased morning activity compared with controls, suggesting possible phase delay.[58] Sleep diary data confirmed decreased sleep efficiency and decreased total sleep in all FTD patients. The disturbances of sleep within the cohort of FTD subjects were less marked than those with moderately severe AD, but changes were present in the FTD group as a whole rather than in only those who had more marked cognitive and behavioural impairment.

Huntington's disease is an inherited neurodegenerative disorder characterized by behavioural and cognitive disturbances and chorea. The sleep phenotype of Huntington's disease includes insomnia, advanced sleep phase, periodic leg movements, RBD, and reduced REM sleep. Reduced REM sleep may precede chorea.[59] Fuh et al. explored the neuropsychiatric manifestations in patients with AD and cortical and subcortical VaD.[60] Patients with cortical VaD had significantly higher mean composite scores in the sleep disturbance domain. This is consistent with a previous small group study showing that VaD patients had more disrupted sleep–wake cycles and decreased sleep quality compared with AD patients.[61] Cheng et al. investigated the association between sleep disturbance, subcortical ischemic vascular dementia, and white matter hyperintensity.[62] They showed that manifestations of sleep disturbance were significantly associated with white matter hyperintensity severity, with most symptoms related to daytime hypersomnolence. In their opinion, disruption of the frontal–subcortical neuronal circuit might play a role in sleep disturbance in patients with subcortical ischemic vascular dementia.

Fatal familial insomnia is an inherited (autosomal dominant) prion disease that was originally described in 1986.[63] It is linked to a mutation at codon 178 of the prion protein gene; it is clinically characterized by a disordered sleep–wake cycle, dysautonomia, dementia and motor signs, and is pathologically characterized by preferential thalamic degeneration.[64] Patients appear apathetic and unable to pay sustained attention to their surroundings. When not stimulated, they tend to become sleepy and manifest recurrent behavioural episodes mimicking a dream (oneiric stupor). The insomnia is not the trivial difficulty in initiating or maintaining sleep, but rather a severe, persistent, and complete disorganization of sleep cycles and a drastic reduction in total sleep time.[65] Death occurs after a disease duration of 7–36 months.[66]

Psychopathology and Sleep Disorders in Dementia

The influence of sleep problems on the behavioural and psychological symptoms of dementia in AD has also been studied. Depressive symptoms are common in older adults, and insomnia is 2.5 times more frequent in older adults with depressed mood.[67] Research has confirmed the association between depression, depressive symptoms, and sleep disorders in persons with dementia.[9] One study concluded that sleep disturbance is a predictive factor of depressive symptoms in AD,[68] and another study revealed an association between apathy and sleep problems in AD.[69] The co-occurrence of sleep complaints in adults without dementia who have anxiety disorders is well known. McCurry et al. provided information about the relationship between anxiety and night-time awakenings in a community-residing sample of individuals with AD. [70] In that study 56% of patients showed symptoms of anxiety, and 29% awakened their caregiver at least once per night during the past week. Moran et al. showed that sleep disturbance in AD is associated with other behavioural symptoms, notably aggressiveness.[71] In line with previous results,[72, 73] García-Alberca et al. demonstrated that sleep disturbances are also significantly associated with daytime behavioural disturbances, namely aberrant motor behaviour and disinhibition.[74] Findings suggest that night-time sleep is important for cognitive function and for behavioural and psychological symptoms of dementia in AD, and that sleep problems should be considered when treating AD patients because their treatment might improve cognitive function as well as behavioural and psychological symptoms of dementia.[75]

Assessment and Management

The goal of managing sleep disturbance in dementia should be to improve both night-time sleep and daytime functioning. Clinical assessment of individuals with sleep disturbances must always include screening for secondary causes, including medical and psychiatric conditions (e.g. depression) and medication side-effects, and for specific sleep disorders. For instance, sleep problems have been related to physical disabilities, respiratory symptoms, cardiovascular problems such as hypertension and cardiac insufficiency,[76] and decreased immune functions.[77] If a patient has a psychiatric disorder or comorbid causes, the disorder or cause should be treated. If sleep disorder is related to medication or drug abuse, the offending medication or drug must be slowly tapered and withdrawn.

An objective baseline measure of a patient's sleep disturbance may be helpful in identifying specific target areas and gauging the efficacy of a proposed intervention. Tractenberg et al. proposed an instrument to assess symptoms of sleep disturbance and disorder, the Sleep Disorders Inventory.[78] It is an expanded version of an item from the Neuropsychiatric Inventory that describes the frequency, severity, and caregiver burden of sleep-disturbed behaviours during a period prior to assessment.[79] The Sleep Disorders Inventory covers a wide range of sleep behaviours and provides information independent of sleep time and sleep quality ratings kept in a diary by caregivers.

Treatment should start with behavioural therapy and, if indicated, should be combined with pharmacologic therapy.

Non-pharmacologic strategies

It seems that behavioural interventions pose a minimal risk of adverse side-effects to patients and should be considered the first line of therapy for sleep disturbance in demented patients. Multifaceted behavioural strategies for improving sleep in patients with dementia have been recommended, but currently there is a paucity of methodologically rigorous research in the area of non-pharmacological sleep interventions for persons with dementia. Many behavioural treatments for insomnia, including stimulus control, sleep restriction, progressive muscle relaxation, biofeedback, sleep hygiene education, paradoxical intention, and multicomponent cognitive-behavioural therapy, are known to be effective with older adults.[80] Sleep hygiene practices to improve sleep include establishing consistent daily times for going to bed and arising from bed, establishing a bedtime routine, and limiting napping to a brief time in the morning or early afternoon.[81]

The American Academy of Sleep Medicine has published practice parameters for the use of bright light to treat sleep and circadian rhythm disorders.[82, 83] Lower levels of light are associated with decreased amplitude of the rest–activity cycle and more wakefulness at night.[84] The goal of light therapy is to expose the patient to increased amounts of natural or artificial light. This added light exposure may provide input to the SCN that will facilitate entrainment of the individual's circadian clock to the 24-h day. Studies examining the efficacy of light therapy have had mixed results. AD patients, whose sleep–wake and rest–activity rhythms are even more severely disrupted, responded well to bright light treatment in many studies.[85, 86] However, no response to bright light therapy in AD patients has been reported by others.[36, 87]

A report showed that acupuncture increases nocturnal melatonin secretion and reduces insomnia and anxiety.[88]

Brown et al. provided a structured critical literature review of the evidence for non-pharmacological interventions to reduce disordered sleep in persons with dementia, and they concluded that there is a paucity of conclusive research for non-pharmacological sleep interventions for persons with dementia.[89] Most of the evidence about effective interventions is anecdotal and untested.

Pharmacologic interventions

There is a full range of medications used to treat sleep disorders, each with particular benefits and potential for harm. Melatonin is synthetized in pineal cells, and its synthesis and release are regulated by seasonal fluctuations in the length of the day. Light exerts an inhibiting effect. Several studies have showed that melatonin levels are diminished in AD patients compared to age-matched control subjects.[90-92] It has been suggested that a deficiency of this hormone in the cerebrospinal fluid is critical for the development of AD because an inadequate concentration of melatonin allows production of free radicalsm which can damage neurons.[93] Singer et al. found that low-dose melatonin (0.5 mg) had a robust effect relative to placebo in two of five AD subjects monitored for a 2-week treatment period by wrist actigraphy,[94] but a follow-up study employing more subjects and a longer monitoring period failed to find such an effect.[95] Large interindividual differences between patients suffering from a neurodegenerative disease are not uncommon and can explain the erratic results seen with melatonin in fully developed AD.[96] It should be also taken into account that melatonin, though having some sedating and sleep latency-reducing properties, does not primarily act as a sleeping pill; it acts mainly as a chronobiotic.[96] Ramelteon, a melatonin receptor agonist, has high affinity for the melatonin (MT) receptors MT1 and MT2. It acts on the MT1 and MT2 receptors to stimulate the action of melatonin to induce and shorten sleep latency.[97] The subjective efficacy of ramelteon was evaluated in clinical trials consisting of 829 elderly outpatients with chronic insomnia; 701 patients (128 patients discontinued) were treated for 5 weeks with 4-mg and 8-mg ramelteon.[98] Patients in both ramelteon groups reported significant reductions in sleep onset latency and increases in total sleep time.

In selected cases, treatment with hypnotics has been useful, but the evidence is limited and care should be undertaken in terms of chronic use, the risk of falls, daytime sedation, and confusion.[99] Benzodiazepines and non-benzodiazepine hypnotics such as zolpidem and zaleplon are frequently prescribed as short-term sleep aids in the general population. Benzodiazepines are associated with increased incidence of sedation, confusion, anterograde amnesia, daytime sleepiness, and rebound insomnia. Even where effective for increasing sleep, benzodiazepines can only be a very short-term solution for occasional use in crises – for example, to give a carer short break from sleep disturbance.[100] Triazolam has a half-life of approximately 2 to 6 h, has clinically significant metabolites, and is rapidly absorbed. However, this medication has been associated with severe rebound insomnia,[101] and its short action may not be optimal for treating early morning awakenings seen in the elderly population.

Antipsychotic medication has often been used for sleep in agitated dementia patients. However, it may aggravate sleep–wake cycle disturbances in AD.[102]

Sleep aids with anticholinergic activity such as tricyclic antidepressants may exacerbate the cholinergic abnormalities inherent to AD and should be avoided.[11] Trazodone offers a dual action on serotonin receptors by blocking serotonergic receptor 2A and inhibiting serotonin reuptake. Its classic indication is for depression, particularly when anxiety and insomnia are also present. It is also widely used for insomnia, although virtually no evidence-based data support its efficacy with older adults.[103] When the risk-benefit ratio of trazodone is assessed, its side-effect profile, which is much more significant than that of conventional hypnotics, should be considered.[104]

Acetylcholinesterase inhibitors are widely used as treatment for AD. There is considerable evidence that the neurotransmitter acetylcholine plays a prominent role in the activation of REM sleep. In addition, there are reciprocal interactions between the cholinergic system and REM facilitatory and inhibitory neurons that maintain the cycle between REM and non-REM sleep.[105] Their cholinomimetic action has the potential to influence sleep quality. In two placebo-controlled clinical trials, donepezil was found to be associated with elevated rates of insomnia based on adverse-event reports.[106, 107] Furthermore, case studies have reported nightmares to be a consequence of donepezil treatment.[108] In contrast, there has been no evidence of night-time sleep-related adverse events in pivotal trials of rivastigmine or galantamine.[109-113] The favourable profile in relation to sleep of galantamine treatment at dosages of 16 or 24 mg/day may be linked to the drug's short half-life, dual mechanisms of action including effects on both muscarinic and nicotinic receptors, and/or timing of administration.[114]

There are currently no clear recommendations for the treatment of sleep impairments in PD with dementia. Positive results have been obtained in the treatment of daytime drowsiness in patients with PD using the selective noradrenaline reuptake inhibitor atomoxetine.[115] Litvinenko et al. evaluated the effects of treatment with galantamine.[116] They found that it produced significant improvements in the quality of nocturnal sleep (restoration of its structure, decreased fragmentation), with reductions in the severity of REM sleep behaviour disorder, daytime drowsiness, and cognitive deficits and hallucinations.


The magnitude and significance of the issue of disordered sleep among elderly individuals in general and, particularly, among those affected by dementia has been clearly demonstrated. The aetiology of sleep disorders is complex, involving multiple factors, such as neurodegenerative changes in the brain, environment, medical or psychiatric morbidity, and medications used to treat chronic illnesses and dementia-related behavioural symptoms. Sleep disturbance occurs in many forms of dementia, but it is possible that certain forms of dementia may be more likely to be associated with disturbed sleep. Additionally, the disordered sleep of dementia patients can compromise the overall health of caregivers, as their sleep is likely to be negatively impacted.

Effective, pragmatic interventions are largely anecdotal and untested. Treatment should target both the sleep problem and any comorbidities in order to optimize the chance for improvement in quality of life and functioning in older adults. There is a need for rigorous scientific inquiry, coupled with tacit knowledge and clinical experience regarding effective interventions, to build strong evidence on the non-pharmacological interventions for disordered sleep for persons with dementia. Pharmacologic options should be used judiciously, with potential side-effects seriously considered before hypnotic and psychotropic agents are prescribed. When drug therapy is used, short-term use is recommended.