Sleep disturbances in people with dementia

There are well-recognized and reported changes to the sleep of PWD over and above the normal age-related changes of shorter, lighter and more easily disturbed sleep (Bliwise et al., Reference Bliwise, Carroll, Lee, Nekich and Dement1993; Bliwise, Reference Bliwise2004). Sleep alterations in dementia can be regarded as an amplification of these normal changes in sleep and they have important sequelae for carers. The deterioration of sleep and circadian rhythmicity in PWD can be seen as part of a continuum, as people with mild cognitive impairment (MCI) have been shown to have worse sleep than age-matched healthy controls, but better sleep than those with a diagnosis of dementia (Beaulieu-Bonneau and Hudon, Reference Beaulieu-Bonneau and Hudon2009). Rongve et al. (Reference Rongve, Boeve and Aarsland2010) reported that 70% of their group of PWD (n = 151) were experiencing some form of sleep disturbance (as reported by their carers), which was worse for people with dementia with Lewy bodies than those with other forms of dementia (Rongve et al., Reference Rongve, Boeve and Aarsland2010).

In addition to these general changes in the circadian regulation of sleep and alterations in the duration and maintenance of sleep, dementia has also been associated with two specific sleep-wake disorders: sundowning and REM sleep Behavior Disorder (RBD). Sundowning in PWD presents as increased levels of wandering and agitation without any specific origin or purpose and has been defined as:

This “exacerbation of symptoms” includes the presence of some or all of the following disruptive behaviors: combativeness; agitated/purposeless movement; wandering; prolonged incoherent vocalization (over five minutes); hallucinations; confusion; and disorientation (Gallagher-Thompson et al., Reference Gallagher-Thompson, Brooks, Bliwise, Leader and Yesavage1992). The deterioration of the circadian rhythm, commonly seen in PWD (Witting et al., Reference Witting, Kwa, Eikelenboom, Mirmiran and Swaab1990; Vitiello et al., Reference Vitiello, Bliwise and Prinz1992; Bliwise et al., Reference Bliwise, Carroll, Lee, Nekich and Dement1993), has been implicated in the phenomenon of sundowning. It has been estimated that between 10% and 25% of the nursing home population display behaviors of the sundowning type (Martin et al., Reference Martin, Marler, Shochat and Ancoli-Israel2000) of which wandering behaviors have been estimated to occur in around 17% of PWD (Klein et al., Reference Klein1999). Wandering has been reported to be more prevalent in males, long-term users of neuroleptic medications, people with Alzheimer's disease (AD), and in older adults with dementia who have experienced an increased duration and severity of the disease. Moreover, sleep disturbances have been implicated as significant predictors of wandering behaviors (Klein et al., Reference Klein1999) and as instrumental in the acceleration of cognitive decline (Prinz et al., Reference Prinz, Vitiello, Bokan, Kukull, Russo and Vitaliano1987; Meguro et al., Reference Meguro1995).

Sometimes PWD experience such a fragmented sleep/wake cycle that agitated behavior, social dysfunction, wandering at night and pacing during the day become a common part of their everyday lives (Ancoli-Israel et al., Reference Ancoli-Israel1997). In extreme cases certain individuals have been seen to be asleep more during the daylight hours than they are at night – the so-called day-night reversal, although complete reversal of 24-hour activity distribution is rare (Allen et al., Reference Allen, Seiler, Stahelen and Speigel1987; Vitiello et al., Reference Vitiello, Bliwise and Prinz1992).

RBD is a parasomnia which is characterized by the loss of REM sleep atonia (a principal feature of REM sleep), i.e. people with RBD lose REM sleep-associated paralysis and can move (sometimes quite violently) during periods of REM and non-REM sleep. People with RBD have been found to display generalized REM and non-REM sleep motor dyscontrol (Schenck et al., Reference Schenk, Bundlie, Ettinger and Mahowald1986). People with RBD commonly display bursts of vigorous and uncontrolled limb movements and vocalization during sleep which is associated with dream recall (Boeve et al., Reference Boeve1998). Losses of brainstem neurons in the locus coereleus and the substantia nigra have been shown to be characteristic of idiopathic RBD in older people (Uchiyama et al., Reference Uchiyama1995). Reported findings have indicated a male predominance of the disease in the order of 87–92% (Schenk et al., Reference Schenck, Hurwitz and Mahowald1993; Boeve et al., Reference Boeve1998). Furthermore, evidence suggests that RBD is symptomatic of dementia with Lewy body disease (LBD) with many people (92%) who display RBD going on to develop dementia with Lewy bodies (DLB) or Parkinson's disease (Uchiyama et al., Reference Uchiyama1995; Boeve et al., Reference Boeve1998; Schenk and Mahowald, Reference Schenck and Mahowald2002). About 15–27% of people with Parkinson's disease have been shown to exhibit RBD, increasing in severity in relation to male gender, high levels of dopaminergic treatment and less severe parkinsonism (Gjerstad et al., Reference Gjerstad, Boeve, Wentzel-Larsen, Aarsland and Larsen2008).

Four types of abnormal circadian rhythms have been identified in PWD (freerunning, phase delayed, ultradian (with rhythmicity of around 3–4 hours), and a flattened amplitude type) (Motohashi et al., Reference Motohashi, Maeda, Wakamatsu, Higuchi and Yuasa2000). These findings indicate the differential nature of circadian rhythm disturbances across PWD and those with different forms of dementing illness, for example, a phase-delayed circadian rhythm has been demonstrated in a group of people with AD, whereas those people with fronto-temporal dementia have been shown to display a relatively unchanged circadian rhythm, but with some indication of phase advancement of their rest activity cycles (Harper et al., Reference Harper, Stopa, McKee, Satlin, Harlan, Goldstein and Volicer2001). People with DLB have been reported to have more severely disturbed circadian rhythms of sleep and wakefulness than those PWD who have AD despite sharing a similar phase delay to their circadian rhythms of core body temperature (Harper et al., Reference Harper, Stopa, McKee, Satlin, Fish and Volicer2004). The inconsistency of circadian rhythm disturbances across dementia subtypes have been reported to implicate dysfunction of the circadian rhythm timing system localized in the suprachiasmatic nucleus (SCN) in the anterior hypothalamic region. The SCN has been shown to be neuronally depleted in PWD, but has also been shown to be relatively unaffected by neurofibrilliary tangles and neuritic plaques (as has the pineal gland) suggesting that, although considerable behavioral and physiological data exist, there remains a limited understanding of the pathophysiology of the circadian timing system in dementia (Bliwise, Reference Bliwise2004). The etiopathogenesis of sleep disturbances in dementia has been reported by Dauvilliers (Reference Dauvilliers2007) who identified abnormal changes to the SCN in PWD, coexisting medical conditions (such as anxiety), sleep-disordered breathing, medication use, genetic predisposition to AD (the presence of the apolipoprotein E ε4 allele), poor sleep hygiene, daytime napping behaviors, and exposure to bright light as contributory factors in sleep and circadian rhythm disorders seen in PWD (Dauvilliers, Reference Dauvilliers2007). Ying-Hui and Swaab (Reference Ying-Hui and Swaab2007) corroborate these observations, but also suggest that a decreased input to the SCN and alterations in the secretion of melatonin by the pineal gland are possible contributors to disturbed circadian rhythmicity in PWD.

Sleep disturbance in carers

Sleep disturbances of PWD are cited by between 46% and 80% of dementia carers as leading to their own sleep disturbances and as a major reason for seeking the permanent institutionalization of their dependents with dementia (Sanford, Reference Sanford1975; Pollak and Perlick, Reference Pollak and Perlick1991; Kesselring et al., Reference Kesselring, Krulik and Bischel2001; Bliwise, Reference Bliwise2004; McCurry et al., Reference McCurry, Logsdon, Teri and Vitiello2007). Caring for someone with dementia precipitates significant decrements to the sleep of carers (Pollak and Stokes, Reference Pollak and Stokes1997; Wilcox and King, Reference Wilcox and King1999) and to carer depression (Livingston et al., Reference Livingston, Manela and Katona1996). Sleep disturbances in carers of PWD have been linked to poorer physical health outcomes independent of carer age; to depression status; to carer self-rated health (Spira et al. Reference Spira2010); and to carer physical and emotional role limitations, energy/vitality, mental health outcomes and health-related quality of life (Lee, Reference Lee2008). Compared with age-matched non-caregiving controls, dementia carers have been reported to sleep less (approximately five hours per night compared to around six hours in controls) and to have reduced sleep efficiencies (around 70% compared to 81% in controls) (Beaudreau et al., Reference Beaudreau2008).

Although sleep problems are common in carers they are often not reported and so remain untreated. In a sample of 119 carers of older PWD, Gilleard (Reference Gilleard1984) reported moderate to significant difficulties with caregiving that were solely associated with nocturnal wandering in 94% of his sample of dementia carers. Gilleard (Reference Gilleard1984) also described, in three separate studies of older adult carers, a mean of 52% of carers complaining about their dependent older adult wandering the house at night (total sample n = 301). Gilhooly (Reference Gilhooley, Hanley and Hodge1984) and Pollak and Perlick (Reference Pollak and Perlick1991) suggested that nocturnal wandering, and the negative impact that this has on the sleep of carers, was a significant predictor of the decision to seek permanent institutionalization for their dependents. Gilhooly (Reference Gilhooley, Hanley and Hodge1984) reported on the feelings of 46 carers of PWD who found that night wandering “was virtually intolerable” (Gilhooly, Reference Gilhooley, Hanley and Hodge1984). Pollak and Perlick (Reference Pollak and Perlick1991) examined reasons reported by 73 community-dwelling dementia carers for their decisions to seek permanent institutional placements for their care recipients. They reported that 80% of these carers experienced sleep disturbances in their care recipients and that 66% of these carers cited sleep disturbance factors as a primary reason for seeking permanent institutional placements for their care recipients (Pollak and Perlick, Reference Pollak and Perlick1991).

Assessment strategies

The assessment of sleep disturbances in PWD and their carers has often been poorly conducted as a result of: (1) lack of recognition of the problem (Rosen et al., Reference Rosen, Rosekind and Rosevear1993); (2) lack of physician training in the assessment of sleep disturbances in these groups (Papp et al., Reference Papp, Penrod and Strohl2002); (3) lack of knowledge about appropriate and available treatment strategies; and (4) the fact that a large number of people live alone (or do not have a bed partner) making it difficult to identify sleeping problems (Ebly et al., Reference Ebly, Hoban and Rockwood1999; Miranda-Castillo et al., Reference Miranda-Castillo, Woods and Orrell2010). Yesavage et al. (Reference Yesavage2003) identified a lack of definitive criteria by which to define sleep disturbances in PWD, impeding the design and delivery of effective research in this area, and so limiting the availability of reliable evidence for the development of various interventions for PWD and their carers. This lack of definitive diagnostic criteria within both DSM-IV and ICD-10 is likely to remain in upcoming updates (DSM-V and ICD-11) as there have been no reported amendments to the current diagnostic criteria for sleep disturbance in dementia, despite planned changes to diagnostic criteria for sleep disorders and dementia in isolation. Indeed, the rebranding of “dementia, delirium, amnestic and other cognitive disorders” into “Major and Minor Neurocognitive Disorders” only mentions sleep once (in relation to delirium) in the newly proposed revisions to DSM-V (Jeste et al., Reference Jeste2010).

Polysomnography (PSG) is regarded as the “gold standard” for sleep assessment, although in vulnerable groups such as PWD and their carers, PSG is not well tolerated, and asking these people to come and have their sleep assessed in a sleep laboratory is confusing for them and does not offer reliable, “naturalistic” sleep assessment. Portable PSG is available to assess sleep in the home environment, but the technology is expensive, requires specialist technicians to operate it, and again, is not well tolerated. Actigraphy offers the opportunity to examine sleep in a naturalistic way, whereby sleep and wakefulness can be measured in people's own homes, using small and unobtrusive devices, over longer periods of time than is possible with PSG (weeks, rather than days) (American Sleep Disorders Association, 1995). These devices and measurement protocols are well tolerated by PWD and their carers (Lee et al., Reference Lee, Morgan and Lindesay2007), and offer an opportunity to explore the sleep and wakefulness patterns of these vulnerable groups, beyond that which is possible with PSG methods, and actigraphy has been specifically recommended for use with PWD (Ancoli-Israel, Reference Ancoli-Israel2005). It should be noted however, that the use of actigraphy in people with Parkinson's disease has been questioned as a result of its inability to distinguish sleep and wakefulness in the presence of disease-associated tremor (Peretz-Lloret et al., Reference Peretz-Lloret, Rossi, Nouzeilles, Trenkwalder, Cardinali and Merello2009).

Other screening measures which may prove useful in the identification of sleep disturbances in PWD and their caregivers include the Pittsburgh Sleep Quality Index (PSQI; Buysse et al., Reference Buysse, Reynolds, Monk, Berman and Kupfer1989) and the Epworth Sleepiness Scale (ESS; Johns, Reference Johns1991); both of these measures are relatively quick to administer and have been used successfully in PWD and their caregivers (Boddy et al., Reference Boddy, Rowan, Lett, O'Brien, McKeith and Burn2007; Lee et al., Reference Lee, Morgan and Lindesay2007, Lee, Reference Lee2008).

Clinical recognition of sleep disturbances in PWD and their carers

Although sleep disturbances in PWD and carers are widely acknowledged in the literature and are seen as indicative features of neurodegenerative disease and carer burden, the recognition of sleep disturbances in these groups often receives sparse attention in diagnostic and treatment procedures. A review by Yesavage et al. (Reference Yesavage2003) identified the following weaknesses in the currently available diagnostic standards for sleep/wake cycle disturbances in people with AD: too broad a definition of “dementia” without consideration of subtypes with their own specific sleep-related phenomena (e.g. RBD in LBD patients); a lack of consideration as to the severity and duration of the disease process; a sole focus on PSG outcomes; and non-specific behavioral outcomes. They proposed the use of updated provisional diagnostic criteria for sleep-wake disturbances in AD patients including:

1. (1) insomnia or excessive daytime sleepiness (EDS) associated with confusional states

2. (2) PSG, actigraphic or sleep log observation detecting two of the following four phenomena:

   1. a. Increased wake time after sleep onset, causing daytime dysfunction;

   2. b. reduced Total Sleep Time (TST) by at least 25% of the patient's pre-morbid TST, or sleeping for fewer than six hours between 9pm and 6am;

   3. c. increasing daytime napping behaviors relative to pre-morbid napping patterns;

   4. d. desynchronization of the sleep/wake rhythm, i.e. reduced night-time and increased daytime sleep.

3. (3) that the sleep disturbance is associated with a diagnosis of AD, and was not present before the onset of dementia;

4. (4) sleep disturbance not explained by differential diagnosis:

   1. a. other medical disorders (e.g. depression, delirium, chronic pain, etc.)

   2. b. other sleep disorders (e.g. PLMS, RLS, apnoea)

   3. c. presence of parasomnias (e.g. RBD – confirmed by PSG or, if not tolerated, by video recordings)

The assessment of sleep disturbances and their treatment may offer opportunities to improve quality of life in PWD and their carers, potentially reduce the rate of cognitive decline and carer burden, and extend extra-institutional, community-based care delivery. Delaying the move from familial to institutional care placements is a better mode of care provision from the perspective of PWD, their carers and institutional care providers (Yesavage et al., Reference Yesavage2003).

Although previous studies have examined subjectively reported sleep of carers, there have been few which have objectively explored carer sleep disturbances. Lee et al. (Reference Lee, Morgan and Lindesay2007) examined the sleep of PWD/carer dyads across a six-week actigraphic sleep measurement protocol examining the impact of respite care on the sleep of PWD and their carers. Findings from this study indicated that respite care periods were associated with significant improvements in carer sleep time (both objective and subjectively rated) and self-reported well-being (Lee et al., Reference Lee, Morgan and Lindesay2007).

Physician training in sleep assessment and the treatment of insomnia

The widespread nature of sleep disturbances in these groups of vulnerable older people, and the significant impact that caring for these groups has on healthcare providers indicates a priority in the training of physicians to assess and treat such disturbances. However, little training is provided to medical students, qualified old age psychiatrists and the allied health professions in this regard. In order to optimize the quality of life and continued community-based care for PWD, improved training for relevant healthcare professions is required. In a large study of clinical psychology training programs, 5,212 program directors were contacted to complete an online survey on sleep education in their program. Only 6% of programs offered formal didactic courses in sleep, while 31% of programs offered training in the treatment of sleep disorders. Only a few programs were linked with a recognized sleep centre (16%), and most reported that their institutions were ineffective in providing sleep education (Meltzer et al., Reference Meltzer, Phillips and Mindell2009). In terms of physician training, only a few studies have examined sleep training. Rosen et al. (Reference Rosen, Rosekind and Rosevear1993) reported that 37% of U.S. medical schools failed to have structured sleep medicine sessions in their curricula; less than 2 hours of teaching time was devoted to sleep; and only 11% of medical students experienced the clinical evaluation of sleep disorders in their training. This lack of training and deficits in knowledge about the treatment of sleep disorders was examined in a U.S. study performed by Papp et al. (Reference Papp, Penrod and Strohl2002). Of 580 primary care physicians surveyed, none rated their knowledge of treating sleep disorders as excellent, 10% as having good knowledge, 60% as fair and 30% poor (Papp et al., Reference Papp, Penrod and Strohl2002). Training in sleep assessment and management is probably little different in other countries.

Benca (Reference Benca2005) reviewed the literature on physician training for insomnia and suggested that several factors contribute to the unmet needs of people with insomnia including: low reporting of insomnia by patients; limited physician training; time constraints; misconceptions about the seriousness of insomnia; a lack of awareness of the advantages of treatment; and ignorance of the potential side-effects of hypnotic use. Non-pharmacologic treatments have been shown to produce reliable and durable changes in people with chronic insomnia, without clinically significant side-effects (Morin et al., Reference Morin, Hauri, Espie, Speilman, Buysee and Bootzin1999); and pharmacological treatments have been shown to be effective with improving: wake time after sleep onset; sleep maintenance; and reducing the number of nighttime awakenings. Pharmacological treatments often precipitate negative side-effects and there is no conclusive evidence at this stage to favor either pharmacological or psychobehavioral treatment of insomnia (Benca, Reference Benca2005; Mallon et al., Reference Mallon, Broman and Hetta2009).

An unmet clinical need

Another clinical problem is that even where sleep disturbance is identified it can be very difficult to determine its clinical importance. Sleep disturbance is highly prevalent in older people generally and, although more frequent in PWD, it is often unclear when intervention is needed. Although effective therapy for sleep disturbance is available it is difficult to apply it appropriately in this patient group or to vulnerable carers (McCurry et al., Reference McCurry, Logsdon, Teri and Vitiello2007). Enabling accurate identification of at-risk groups and better sleep assessment should improve the use of appropriate treatment strategies for them.

Although sleep and activity intervention studies in these groups are sparse, there is evidence that the manipulation of patient/carer environments and/or behaviors can have a positive impact on the sleep experienced by these groups (Hinchcliffe et al., Reference Hinchcliffe, Hyman and Blizard1992; King et al., Reference King, Baumann, O'Sullivan, Wilcox and Castro2002).

Management of sleep problems in PWD

The management of sleep problems in PWD has been implemented with both pharmacological and non-pharmacological (psycho-behavioral) methods. Sleep problems can be effectively treated in older people with or without dementia using a stepped care approach (Ancoli-Israel, Reference Ancoli-Israel2005). In those with mild disturbance, a medication review to identify and consider withdrawing or switching culprit medication associated with insomnia (e.g. B-blockers) should be conducted. This includes “over the counter” drugs such as decongestants and non-prescribed substances (alcohol, caffeine, nicotine). Simple advice on sleep hygiene and the use of information leaflets are also appropriate. In moderate sleep disturbance (or where these measures are ineffective) more formal sleep hygiene education supplemented by techniques to resynchronize disturbed circadian rhythms in older people (including for those with dementia) can be used (e.g. light therapy, melatonin, and increased physical activity) (Ying-Hui and Swaab, Reference Ying-Hui and Swaab2007). At the next level is cognitive behavioral therapy, which can be effective in abbreviated form (Edinger and Sampson, Reference Edinger and Sampson2003) and used in primary care, or for PWD, psychobehavioral treatments (e.g. sleep hygiene, sleep restriction, stimulus control) (McCurry et al., Reference McCurry, Logsdon, Teri and Vitiello2007) are as effective as drug interventions for treating sleep disturbance (Jacobs et al., Reference Jacobs, Pace-Schott, Stickgold and Otto2004).

Finally, in selected cases with severe sleep disturbance, hypnotic drugs may be used, after giving careful consideration to potential side-effects (Wilson and Nutt, 2008). These are effective in PWD and can avoid the use of antipsychotics which are associated with potentially more severe adverse effects. Evidence for the efficacy of both these approaches is limited (Sink et al., Reference Sink, Holden and Yaffe2005; McCurry et al., Reference McCurry, Logsdon, Teri and Vitiello2007; Spjiker et al., 2008; Banerjee, Reference Banerjee2009). Nevertheless, sleep problems have been effectively treated in PWD and their carers with pharmacological (e.g. benzodiazepines, newer “z” drugs such as zopiclone, and, more recently, melatonin (Wilson and Nutt, 2008)), or with psychobehavioral treatment modalities (e.g. sleep hygiene, sleep restriction, stimulus control and cognitive behavioral therapy for insomnia – CBTi (McCurry et al., Reference McCurry, Gibbons, Logsdon, Vitiello and Teri2005)). Other treatment modalities for the treatment of sleeping problems in older adults include antihistamines and herbal preparations, although no research is available to suggest the safety or efficacy of these over-the-counter agents (Ancoli-Israel, Reference Ancoli-Israel2005).

Management of sleep disturbance in carers of PWD

A similar stepped care approach to the treatment of sleep disturbance in PWD is recommended for the management of sleep disturbance among carers. Treatment again falls into two the broad categories of pharmacological and non-pharmacological treatments and these will be reviewed in turn below.

Conflict of interest

None.

Description of authors' roles

Dr. Lee researched, prepared and wrote the paper. Dr. Thomas commented on drafts and contributed to writing the paper.