Introduction
Sleep problems in older adults are common: up to 42% of people over 65 report problems in initiating or maintaining sleep.Reference Foley, Monjan, Brown, Simonsick, Wallace and Blazer1 Problems with sleep in older people have been comprehensively reviewed elsewhereReference Ancoli-Israel and Ayalon2 and are due to medical problems,Reference Katz and McHorney3–Reference Quan, Katz, Olson, Bonekat, Enright, Young and Newman6 an increased prevalence of specific sleep disordersReference Ohayon and Roth7, Reference Ancoli-Israel, Kripke, Klauber, Mason, Fell and Kaplan8 and age-related changes in circadian rhythm.Reference Bliwise9 Care home residents are, however, an extreme case – they have a higher prevalence of frailty and cognitive impairment than their community-dwelling counterparts,Reference Bowman, Whistler and Ellerby10 have lower levels of physical activity,Reference MacRae, Schnelle, Simmons and Ouslander11 have daily routines altered to fit with institutional timetables and spend much of their time indoors.Reference Shochat, Martin, Marler and Ancoli-Israel12 All of these variables may influence circadian rhythm.
In this article we consider sleep problems specifically in the care home population and present findings from randomized controlled trials (RCTs) of interventions undertaken to improve the quality of sleep in this cohort, including evaluations of pharmacological, physical, sleep hygiene, light and combination therapies. We also consider implications for practice and future research.
Search strategy
This work comprises part of a larger literature review of RCTs conducted in the care home setting. Three databases were consulted, with different search terms and limits used for each based upon those available. Medline (1950–June 2009) was searched for the search terms ‘nursing home’, ‘residential facilities’ and ‘homes for the aged’, combined using the ‘OR’ command. Results were limited for English language and randomized controlled trials. The Allied and Complementary Medicine Database (AMED) (1985–June 2009) was searched for ‘nursing homes’, ‘long term care’ and ‘residential facilities’ combined using the ‘OR’ command and ‘randomized controlled trial’ using the ‘AND’ command. The British Nursing Index and Archive (BNI) (1985–June 2009) was searched for ‘nursing homes’, ‘residential care’ and ‘long-term care’. 563 articles were identified from Medline and AMED and 3161 from BNI – it was not possible to limit the results from this database further as no filter for RCTs was available. Abstracts were reviewed by a single reviewer for descriptions of interventions evaluated using RCTs in residential, nursing or care homes. 339 articles met these criteria and were read in full. From these, articles reviewing interventions for sleep in care home residents were used to compile the current review.
What is a care home?
Long-term care facilities are given different titles depending on the country in which they operate. A care home is defined in the United Kingdom as ‘an establishment [which] provides accommodation, together with nursing or personal care, for persons who are or have been ill, who have or have had a mental disorder, who are disabled or infirm, or are or have been dependent on alcohol or drugs’.13 The majority of homes care for residents with frailty and physical dependency, with those caring for people with learning difficulties, alcohol or drug dependency and mental health problems separated by specialist registration. All care homes provide support with activities of daily living but are classified as either residential care homes or care homes with nursing depending on whether they provide dedicated 24-hour professional nursing care. When the sector is considered as a whole, only 8.6% of residents are under 70, with 76% overall requiring assistance with their mobility or being immobile, 78% having at least one form of mental impairment and 71% suffering incontinence.Reference Bowman, Whistler and Ellerby10 Although most developed countries have a care home equivalent, the model of separate residential and nursing care is far from universal.Reference Ribbe, Ljunggren, Steel, Topinková, Hawes, Ikegami, Henrard and Jónnson14 Despite these differences, we present in this article findings that we believe are applicable across national boundaries.
The problem of sleep in care homes
Sleep is often fragmented in care home residents. A study of 200 residents, conducted in 1989, gathered recordings of sleep–wake cycles using a portable Respitrace-Medilog device.Reference Ancoli-Israel, Parker, Sinaee, Fell and Kripke15 This was carried by the participant, and used thoracic and abdominal inductance bands to measure respiration, tibialis electromyography to record hypnogogic leg myoclonus and a wrist transducer to measure arm and hand activity. Recordings averaging 15.4/24 hours per subject were obtained, and revealed that subjects averaged a relatively normal 7 hours 58 minutes asleep and 7 hours 28 minutes awake. However, in the sleep periods, they spent no more than 39.5 minutes per hour asleep, whilst 50% of subjects woke up more than twice per hour. It was noted that residents compensated by catnapping during normal waking hours.
There are several reasons for disturbed sleep patterns amongst care home residents. They are, for the most part, old and older people experience a higher prevalence of sleep disorders. This is attributed to an accumulation of general medical conditions associated with insomnia (heart failure,Reference Katz and McHorney3 Parkinson's disease,Reference Garcia-Borreguero, Larrosa and Bravo4 dementia,Reference Vitiello, Prinz, Williams, Frommlet and Ries5 depressionReference Quan, Katz, Olson, Bonekat, Enright, Young and Newman6), a higher prevalence of sleep-specific disorders such as sleep apnoeaReference Ancoli-Israel, Kripke, Klauber, Mason, Fell and Kaplan8 and restless leg syndrome,Reference Ohayon and Roth7 and age-associated changes in the sleep–wake cycle that result in early morning wakening and difficulty maintaining sleep. The latter are attributed in part to degenerative changes in the suprachiasmatic nucleus of the hypothalamus, commonly referred to as the pacemaker of circadian rhythm.Reference Ancoli-Israel and Ayalon2
Circadian rhythms are set by ‘zeitgebers’ (from German, literally meaning ‘time givers’) – including bright light and physical activity – to which care home residents have limited exposure. Bright light exposure during the day, of 2000 lux or more, helps maintain circadian rhythm. Yet data recorded using wrist-worn photosensors from 66 care home residents recorded a median exposure of only 10 minutes per day of light above 1000 lux intensityReference Shochat, Martin, Marler and Ancoli-Israel12 compared with average exposures amongst young adults of 58 minutes per dayReference Espiritu, Kripke, Ancoli-Israel, Mowen, Mason, Fell, Klauber and Kaplan16 and healthy older people of 60 minutes per dayReference Campbell, Kripke, Gillin and Hrubovcak17 above 2000 lux. Activity also helps maintain circadian rhythm. Yet data gathered from 95 care home residents using motion sensors to measure activity levels found that unrestrained residents spent most of their time immobile – 83.5% of measurements recorded participants sitting or lying flat.Reference MacRae, Schnelle, Simmons and Ouslander11
Regimens designed to facilitate care delivery in care homes, such as night-time continence checks, disturb residents' sleep. A study of 100 residents across four nursing homes, using bedside monitors to record night-time noise and light levels, recorded 32 noises per resident per night at the volume of loud speech (60 decibels) or louder.Reference Schnelle, Ouslander, Simmons, Alessi and Gravel18 Another study using similar technology in 225 residents from 10 homes revealed 22% of waking episodes to be associated with noise alone, 10% with light or light and noise, and 10% with incontinence care routines. Seventy-six per cent of all incontinence care practices resulted in awakenings.Reference Cruise, Schnelle, Alessi, Simmons and Ouslander19
Disturbed sleep is detrimental. A survey of 1526 64- to 99-year-olds recording falls, health and socio-economic variables, demonstrated a strong positive correlation between falls and night-time sleep problems.Reference Brassington, King and Bliwise20 Focusing specifically on care home residents, a retrospective case note analysis of 507 deaths in 1557 residents established standardized mortality rates for commonly presenting problems and showed a hazard ratio of 1.9 (95% CI 1.3–2.8) for excess day time sleepiness (the correlate of night-time sleep disruption).Reference Dale, Burns, Panter and Morris21 This finding is supported by a prospective study of 272 Japanese long-term geriatric hospital patients (the equivalent of UK care home residentsReference Ribbe, Ljunggren, Steel, Topinková, Hawes, Ikegami, Henrard and Jónnson14), which recorded sleep-patterns over 2 weeks using 2-hourly observations at baseline and showed a hazard ratio at 2-year follow-up of 1.6 (95% CI 1.5–2.4) for those with night-time insomnia and 1.8 (95% CI 1.2–2.8) for those with sleep-onset delay.Reference Manabe, Matsui, Yamaya, Sato-Nakagawa, Okamura, Arai and Sasaki22
Given that sleep quality is so poor amongst care home residents and that this is likely to be detrimental to their health, it is unsurprising that research has focused on treatments to improve the quality and quantity of sleep in this cohort.
Pharmaceutical interventions
Hypnotics have been shown to increase sleep duration in the older populationReference Glass, Lanctot, Herrmann, Sproule and Busto23 and, despite their considerable side-effect profile, they are still recommended by some expert authorities for treatment of short-term insomnia in this age group.Reference Bloom, Ahmed, Alessi, Ancoli-Israel, Buysse, Kryger, Phillips, Thorpy, Vitiello and Zee24 It is likely, however, that they play little role in treating care home residents with insomnia. A randomized controlled crossover trial comparing temazepam with diphenhydramine – a first generation antihistamine available over the counter – and placebo in 14 care home residents reported reduced sleep latency (time between going to bed and sleep onset) for both hypnotics, but no advantage in terms of duration of sleep, number of awakenings or time spent awake.Reference Meuleman, Nelson and Clark25 In addition, participants scored higher in terms of cognitive function testing (word lists, cancellation tests, digit span and digit symbol substitution) whilst receiving placebo – indicating significant hangover effects from both drugs. These findings are supported by a cohort study of 145 US nursing home residents recording baseline medication use and self-reported sleep complaints. There was no demonstrated association between hypnotic prescribing and presence/absence of sleep complaints at either baseline or 6 month follow-up.Reference Monane, Glynn and Avorn26 Therefore, even before the considerable side-effect profile of hypnoticsReference Ray, Thapa and Gideon27 in frail older people is considered, there is little to justify their widespread use. This has led to a search for alternative pharmacological treatments, with specific focus on melatonin and the melatonin-receptor agonist, ramelteon.Reference Roth, Stubbs and Walsh28 Of these, only melatonin has been tested in the care home population.
Melatonin is an endogenous hormone produced in the pineal gland, with its secretion mediated by stimulation from the suprachiasmatic nucleus in response to light. A nocturnal peak in melatonin secretion is clearly demonstrable in young subjects and is thought to play a role in initiating sleep.Reference Zeitzer, Daniels, Duffy, Klerman, Shanahan, Dijk and Czeisler29 Older people have both lower levels of circulating melatonin – measured through serum levels and excretion of serum metabolites – and a diminished nocturnal peak.Reference Magri, Sarra, Cinchetti, Guazzoni, Fioravanti, Cravello and Ferrari30, Reference Skene and Swaab31 Exogenous melatonin has been shown to be effective in promoting shifts in circadian rhythms,Reference Wyatt, Dijk, Ritz-de Cecco, Ronda and Czeisler32 leading to the hypothesis that it may play a role in normalizing sleep–wake cycles in older people with insomnia.
Only one RCT has evaluated melatonin as a stand-alone therapy in the care home population. This selected 41 residents with Alzheimer's disease (National Institute of Neurological and Communicative Disorders and Stroke criteria) and randomized them to receive either melatonin or placebo for 10 days. Sleep efficacy (the proportion of the time in bed spent asleep) was measured using actigraphy, where motion and light sensors are worn in a wrist-watch type device – a method validated against EEG recordings for ascertainment of sleep in care home residents.Reference Ancoli-Israel, Clopton, Klauber, Fell and Mason33 Agitation was recorded using the Cohen Mansfield Agitation Inventory (CMAI) immediately before and after treatment. No significant difference was found for either outcome measure. Although at odds with findings from small cohort studies,Reference Cardinali, Brusco, Liberczuk and Furio34, Reference Cohen-Mansfield, Garfinkel and Lipson35 these findings are supported by a non-care home placebo-controlled trial of melatonin in 157 older patients with Alzheimer's disease, where no effect on sleep efficacy was demonstrated.Reference Singer, Tractenberg, Kaye, Schafer, Gamst, Grundman, Thomas and Thal36
The authors of these studies have queried whether the negative results were a consequence of either the supra-physiological doses of melatonin employed or the phase-shift in circadian rhythm prominent in Alzheimer's disease. It might, for example, be necessary to administer melatonin at a different time of day than in the non-demented population to promote improved sleep efficacy. There remains, however, no evidence to support melatonin as a stand-alone therapy in care home residents. However, there is evidence, as discussed later in the article, to promote its use as part of a multi-component intervention.
Turning up the light
Light is a potent zeitgeber and, given that care home residents are relatively light deprived, there has been considerable interest on how light therapy affects sleep.
Bright light therapy, using electric light boxes, affects circadian rhythm in older people. Evening bright light delays sleep onset such that people fall asleep later and morning bright light advances the onset, leading to earlier onset of sleep.Reference Klerman, Duffy, Dijk and Czeisler37 Non-demented older people characteristically tend to fall asleep earlier than normal controlsReference Carrier, Monk, Reynolds, Buysse and Kupfer38 and therefore potentially could benefit from evening bright light. People with Alzheimer's disease tend to fall asleep laterReference Satlin, Volicer, Stopa and Harper39 and could potentially benefit from morning light.
There have been two RCTs of light as a stand-alone therapy in a care home setting. In the first of these,Reference Ancoli-Israel, Martin, Kripke, Marler and Klauber40 77 care home residents with a mean MMSE of 12.8 were randomized to receive either evening bright light therapy, morning bright light therapy, daytime sleep restriction – where residents were kept awake by study staff – or evening dim light. Evening dim light was the control intervention. Light therapies were delivered using medical-grade light boxes, with bright light at 2500 lux intensity and dim light at 50 lux. Day- and night-time activity levels were measured using actigraphy. No difference between groups was detected and there was no within-group difference in activity between baseline and follow-up. A non-significant tendency for morning bright light to delay the onset of sleep was noted.
The second study selected 46 residents with Alzheimer's disease (mean MMSE 6.7), who were randomized to receive either morning bright light or normal light exposure.Reference Dowling, Hubbard, Mastick, Luxenberg, Burr and Van Someren41 A naturalistic approach was adopted by encouraging residents in the intervention arm to undertake activities either outdoors or in a bright room, with light levels recorded to ensure adequate exposure. When light levels were less than 2500 lux for an hour or more, top-up treatment was provided using light boxes. Sleep efficacy and day and night activity levels were recorded by actigraphy. There was no difference between groups when the intervention cohort was considered as a whole, but positive effects were noted on a subgroup analysis of those individuals (n = 13) who had the most profound sleep disturbance at baseline, with their 10 most active hours occurring at a time they would be expected to be sleeping. For this group there was a statistically significant improvement in rest–activity ratios, night-time sleep efficacy and sleep duration by comparison with controls.
Thus light therapy has little evidence to support its use in the care home population as a stand-alone therapy, except perhaps in patients with dementia and profound sleep–wake cycle disruption.
Multicomponent interventions
It is unsurprising that stand-alone therapies, either melatonin or light, have little influence over sleep or circadian rhythms in the care home population, given that the causes are likely to be multifactorial. Furthermore, the suprachiasmatic nucleus undergoes degenerative change in advanced old age and this is particularly marked in Alzheimer's disease.Reference Swaab, Fliers and Partiman42 It has therefore been proposed that older patients, particularly those with Alzheimer's disease, will be less sensitive to individual zeitgebers such as light exposure or circulating melatonin levels. A response has been to combine individual therapies in the hope that their influence on sleep is summative – melatonin, for example, might prime an individual to respond to bright light therapy.Reference Dowling, Burr, Van Someren, Hubbard, Luxenberg, Mastick and Cooper43
Three multicomponent interventions have been described in the care home setting and have, for the most part, been more successful than stand-alone therapies.
Combination 1 – physical exercise and sleep hygiene
The mechanism of the association between physical exercise and sleep is poorly understood but is likely to be mediated through an effect on overall physical fitness rather than an immediate effect of the exercise, with improved physical fitness resulting in improved quality of sleep.Reference Driver and Taylor44
Functional incidental training is incorporated into daily care home routines by encouraging participants to undertake repetitions of routine activities such as sit-to-stand transfers following toileting and therefore has intuitive advantage over potentially more artificial, structured exercise programmes. It has been shown to improve physical endurance in an RCT of 76 nursing home residentsReference Schnelle, MacRae, Ouslander, Simmons and Nitta45 and hence, putatively, might play a role in improving sleep.
A post-hoc analysis of two RCTs, one involving functional incidental training and the other involving a programme of rowing, walking or wheelchair-propulsion-based exercises, showed no benefit for either intervention on sleep.Reference Alessi, Schnelle, MacRae, Ouslander, al-Samarrai, Simmons and Traub46 The authors attributed this to the possible disruptive effects of light and noise associated with routine night-time checks on residents. A subsequent study allowed for this by combining a 14-week functional incidental training programme with 1 week of sleep hygiene in the form of a ‘quiet at night programme’.Reference Alessi, Schnelle, MacRae, Ouslander, al-Samarrai, Simmons and Traub46 In this staff were advised to wear quiet shoes, minimize entries into residents' rooms, use a flash-light rather than overhead lighting when doing so and not to awaken a patient for continence checks unless they had been asleep for four consecutive one-hourly reviews. Twenty-nine participants were randomized to receive either functional incidental training plus quiet at night, or the quiet at night intervention alone. Sleep efficacy, the proportion of daytime spent asleep and agitation levels all improved in the intervention arm. However, the study was noted to be inconclusive due to technical issues, most importantly the small sample size and inadequate randomization, which led to the intervention subjects having markedly poorer sleep at baseline.
Combination 2 – physical exercise, sleep hygiene and light
In a further augmentation of their protocol and in an attempt to address the sample-size issue these authors conducted a second study evaluating functional incidental training, this time combined with both sleep hygiene and light exposure.Reference Alessi, Martin and Webber47 The intervention comprised 5 days of daytime functional incidental training, coupled to a quiet at night programme and 30 minutes per day of bright outdoor light exposure, confirmed using a handheld meter as being 30 minutes at ≥ 10,000 lux. Control subjects received usual care. 118 participants were individually randomized. Actigraphy recordings demonstrated no effect on the duration of sleep or number of night-time awakenings; however, the duration of night-time awakenings was significantly, if slightly, shorter. A more significant finding was an 11% reduction in the proportion of the day spent asleep and a resultant improvement in rest–activity ratios.Reference Martin, Marler, Harker, Josephson and Alessi48 Two specific issues draw this analysis into question – residents were randomized individually, rather than by home, with the possibility that those receiving usual care were partially exposed to the intervention through cross-contamination; and the possibility that social interaction alone acts as a zeitgeber could not be excluded by the use of a usual care control.
Combination 3 – melatonin and light
The combination of melatonin and bright light has also been tested.Reference Dowling, Burr, Van Someren, Hubbard, Luxenberg, Mastick and Cooper43 Fifty patients were randomized to receive bright morning light and evening oral melatonin, bright morning light and evening oral placebo or usual care, with sleep recorded using actigraphy. When compared with controls, those receiving both light and melatonin showed higher daytime activity levels, fewer and a shorter duration of daytime sleep episodes and a more normal sleep–wake ratio. There was, however, no effect on night-time sleep. Light and placebo had no effect on sleep; however, the small sample size meant that the study was underpowered to detect such an effect and it is possible that a larger study may have revealed a positive, if smaller magnitude, effect for light alone.
Conclusions
We have summarized in this article the RCT-generated evidence base for sleep interventions in care homes. There is little to suggest that hypnotic drugs, or the other stand-alone therapies that have been tested – bright light therapy and oral melatonin – have any overall benefit. Combination therapies seem to be more effective with functional incidental training plus sleep hygiene, functional incidental training plus light plus sleep hygiene and melatonin plus light all having positive effects. These effects are seen predominantly during the day – with residents staying awake for longer and being more physically active whilst doing so. This improvement in daytime arousal may be positive by allowing improved participation in daytime activities and improving quality of life measures as a consequence.Reference Cohen-Mansfield, Garfinkel and Lipson35, Reference Dowling, Burr, Van Someren, Hubbard, Luxenberg, Mastick and Cooper43 What is not clear, however, is whether it has any meaningful effect on the more dangerous associations of impaired sleeping – falls and mortality. It is not, in fact, clear what the causal relationship is between sleep and these variables and whether sleep interventions, even if effective, can have any influence over them.
It is difficult to make any concrete recommendations based upon current research in care homes because of technical issues with the RCTs undertaken. All are small and have used individual rather than cluster randomization, with implications for cross-contamination. Several studies have issues with poorly chosen control interventions that make it difficult to interpret the true meaning of their results. Also, whilst it is clear that older patients with and without dementia have differing circadian rhythm disturbances, it is not clear that these groups have been considered separately in several of the studies undertaken.
Further research is required if proper sense is to be made of how to manage sleep in care homes. Nevertheless, there are practical lessons to note from what is already known. There is no good evidence that there are any useful drug therapies for sleep disturbance in this group and reasons to believe that they are largely ineffective. The evidence that non-pharmacological interventions are effective is not strong either. Stand-alone interventions do not seem to work, so we propose that a sleep strategy must be comprehensive and tackle more than one issue at once. Physical activity seems to have a positive influence on sleep and this effect can be demonstrated with relatively low levels of exertion that could easily be achieved as part of a care home activity programme. Light exposure seems to help patients sleep and the examples from the research of this being achieved by outdoor sun exposure could be easily delivered as part of care home routines. Finally, attempts to reduce night-time disruption are important. This latter concept will trouble many care home practitioners because of the tension between providing good pressure care through regular continence checks and leaving patients to sleep. The optimum regimen that allows both issues to be safely addressed is yet to be established and is a potential focus for future research.
Conflicts of interest
The authors declare that they have no conflicts of interest.
