International Psychogeriatrics | My Assignment Tutor

International Psychogeriatrics (2019), 31:2, 251–265 © International Psychogeriatric Association 2018. This is an Open Access article, distributed under the terms of the CreativeCommons Attribution licence (, which permits unrestricted re-use, distribution, and reproduction in any medium, provided the originalwork is properly cited.doi:10.1017/S1041610218000753DREAMS-START (Dementia RElAted Manual for Sleep;STrAtegies for RelaTives) for people with dementia and sleepdisturbances: a single-blind feasibility and acceptabilityrandomized controlled trial…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………….Gill Livingston,1,2 Julie A. Barber,3 Kirsi M. Kinnunen,1 Lucy Webster,1 Simon D. Kyle,4Claudia Cooper,1,2 Colin A. Espie,4 Brendan Hallam,1 Rossana Horsley,5James Pickett6 and Penny Rapaport11Division of Psychiatry, Faculty of Brain Sciences, UCL, London, UK2Services for Ageing and Mental Health, Camden and Islington NHS Foundation Trust, London, UK3Department of Statistical Science, Faculty of Mathematical & Physical Sciences, UCL, London, UK4Sleep and Circadian Neuroscience Institute (SCNi), Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, UK5Alzheimer’s Society Research Network, London, UK6Alzheimer’s Society, London, UKABSTRACTBackground: 40% of people with dementia have disturbed sleep but there are currently no known effectivetreatments. Studies of sleep hygiene and light therapy have not been powered to indicate feasibility andacceptability and have shown 40–50% retention. We tested the feasibility and acceptability of a six-sessionmanualized evidence-based non-pharmacological therapy; Dementia RElAted Manual for Sleep; STrAtegiesfor RelaTives (DREAMS-START) for sleep disturbance in people with dementia.Methods: We conducted a parallel, two-armed, single-blind randomized trial and randomized 2:1 tointervention: Treatment as Usual. Eligible participants had dementia and sleep disturbances (scoring ≥4 onone Sleep Disorders Inventory item) and a family carer and were recruited from two London memory servicesand Join Dementia Research. Participants wore an actiwatch for two weeks pre-randomization. Trained,clinically supervised psychology graduates delivered DREAMS-START to carers randomized to intervention;covering Understanding sleep and dementia; Making a plan (incorporating actiwatch information, lightexposure using a light box); Daytime activity and routine; Difficult night-time behaviors; Taking care ofyour own (carer’s) sleep; and What works? Strategies for the future. Carers kept their manual, light box,and relaxation recordings post-intervention. Outcome assessment was masked to allocation. The co-primaryoutcomes were feasibility (≥50% eligible people consenting to the study) and acceptability (≥75% ofintervention group attending ≥4 intervention sessions).Results: In total, 63out of 95 (66%; 95% CI: 56–76%) eligible referrals consented between 04/08/2016 and24/03/2017; 62 (65%; 95% CI: 55–75%) were randomized, and 37 out of 42 (88%; 95% CI: 75–96%)adhered to the intervention.Conclusions: DREAM-START for sleep disorders in dementia is feasible and acceptable.Key words: sleep disorders, carers, cognitive behavioral therapy (CBT), randomized controlled trial (RCT), physical activityIntroductionCurrently, 47 million people live with dementiaworldwide, with numbers expected to nearly tripleCorrespondence should be addressed to: Gill Livingston, Division of Psychiatry,Faculty of Brain Sciences, UCL, Maple House, 149 Tottenham CourtRoad, W1T 7NF London, UK. Phone: 00 44 207 679 9435. Received 5 Mar 2018; revision requested 12 Apr 2018;revised version received 13 Apr 2018; accepted 17 Apr 2018. First publishedonline 17 September 2018.Trial registration: The full protocol can be accessed 2050 because of increasing longevity (Princeet al., 2015). Sleep disturbances are commonin dementia, for example, occurring in around40% of people with Alzheimer’s disease (AD)(Moran et al., 2005; Dauvilliers, 2007; Zhao,2016). Causes of sleep disturbance are variedand include pain relating to comorbid disorders,disorientation to time of day, or neuropsychiatricsymptoms, including anxiety and depression. Inaddition, dementia may impair the sleep–wake from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at252 G. Livingston et al.cycle through degeneration of the suprachiasmaticnucleus, disrupting the normal circadian rhythm(Ju et al., 2014). Sleep disruption reduces everydayfunction and quality of life (Kyle et al., 2010).It also increases family carer burden, predictsdepressive symptoms, and leads to care homeadmissions, thus increasing individual, societal, andeconomic costs associated with dementia (McCraeet al., 2016; Livingston et al., 2017).There are, however, currently no knowneffective treatments for sleep problems in dementia,although there have been trials of drugs, includingmirtazapine and melatonin which have beenineffective (McCleery et al., 2014; Livingston et al.,2017). Other studies incorporating sleep hygieneand light therapy have been too small for definitiveresults (McCurry et al., 2009; Forbes et al., 2014;Kinnunen et al., 2017). Health teams use a mixtureof sleep hygiene measures and psychotropic medication, extrapolated from other conditions, thatgive limited benefit, and medications may have sideeffects. Bright light therapy to strengthen circadianrhythmicity has some effect on sleep disturbancesin the general population (McCurry and AncoliIsrael, 2003). Cognitive behavioral techniques forsleep management have been effective in olderadults without dementia and in family carers ofpeople with dementia (Montgomery and Dennis,2003; Sivertsen and Nordhus, 2007).The objective of this study is to test the feasibilityand acceptability of Dementia RElAted Manualfor Sleep; STrAtegies for RelaTives (DREAMSSTART), a multicomponent manualized intervention built on the above available evidence, tomanage clinically significant sleep disturbance inpeople with dementia living in their own homes.MethodsStudy design and participantsThis is a two-armed randomized controlled trial,recruiting from three UK sites; Camden andIslington NHS Foundation Trust, Barnet, Enfieldand Haringey Mental Health NHS Trust, andJoin Dementia Research (JDR), where peopleregister their interest in participating in dementiaresearch. London – Queen Square Research EthicsCommittee (Reference: 16/LO/0670) approvedthe study. We obtained written consent fromall participating family carers and patients withthe mental capacity to give informed consent toparticipation in the trial prior to their enrollment. Ifthe person with dementia did not have capacity togive consent, we required a consultee’s declaration.The trial steering committee provided overallsupervision of the trial with an independentchairperson leading it. The full protocol is in theAppendix (see online supplementary material).Eligible patients from the trusts lived in theirown home, had a clinical diagnosis of dementia anda sleep disorders inventory (SDI) item score ≥4(Tractenberg et al., 2003), judged as problematicby the person with dementia or their family,and had a primary family carer who providedsupport at least weekly. People with dementia,who had a primary sleep disorder diagnosis, suchas sleep apnoea, were excluded. Patients andcarers were assessed at home after consent andat follow-up three months later. The baselinemeasures included socio-demographic details anddementia severity through the Clinical DementiaRating (CDR), a reliable and valid instrument forclassifying clinically diagnosed dementia (Hugheset al., 1982; Morris, 1997) as very mild (0.5), mild(1), moderate (2) or severe (3). After enrollment,we asked patients to wear wrist-worn actiwatches(MotionWatch 8) CamNtech Ltd. 24 h a dayfor two weeks before randomization and again atfollow-up.Randomization and maskingAn independent statistician produced computergenerated randomization lists stratified by site andbased on random permuted blocks of sizes threeand six to allow 2:1 allocation to intervention:Treatment as Usual (TAU) groups. Therapistsworked in two separate teams of two therapists eachand assessed outcomes for participants, to whomthe other team had delivered the intervention,masked to group allocation. A third non-therapistresearcher assessing outcomes was masked to allallocations. Due to the intervention’s nature it wasnot possible to mask the trial participants. Whenarranging follow-up, participants were remindednot to disclose their allocation group to theassessor, and to remove from view anything relatedto DREAMS-START.ProceduresThose allocated to the intervention received the sixsession DREAMS-START manual. We developedDREAMS-START for the study in an iterative coproduction process, involving dementia and sleepexperts and Public and Patient Input (PPI) (PR,GL, SK, CE, CC with PPI led by RH and JP).We incorporated prior evidence, some existingmaterials and used the START manual structure(Livingston et al., 2013; Livingston et al., 2014).This iterative coproduction is reported in detailelsewhere (Kinnunen et al., 2017). DREAMSSTART comprised cognitive-behavioral components, including psychoeducation, light therapy,establishing a new sleep–wake schedule (based on from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available atDREAMS-START to improve sleep in dementia 253Six sessions comprising:1. Understanding sleep and demena: Psychoeducaon on the importance of sleep, sleepprocess, the impact of demena upon sleep, lifestyle and bedroom environment factorsthat could affect sleep, the impact that sleep problems can have upon the person withdemena and their relave.2. Making a plan: Plan for the person with demena to increase natural and arficial light;standardise meal mes; changes to bed and rise mes; reduce dayme naps (plan basedupon the individual’s acgraphy data and sleep diary). Establishing adapve smuluscontrol: pre-bed seling roune and management of wakeful episodes. The carers weregiven a ‘light box’ (Lumie Arabica SAD Light, Lumie, Cambridge, UK) to keep and it was leon during the sessions to help people become habituated to the bright light(10,000 lux at 25cm). We provided a me switch if carers wanted it andrecommended it was le on for 30 minutes at the same me every morning.3. Dayme acvity and roune: Promong de-arousal at night (e.g. relaxaon, bedroomcomfort, no caffeine or alcohol intake before bed, no acvies in bedroom) andbehavioural acvaon during the day to maintain alertness, reduce dayme naps. Buildingpleasant acvies and exercise/physical acvity into the day, including a seated exercisevideo for less physically able individuals. Every individual made an exercise and acvityplan.4. Difficult night-me behaviours: Troubleshoong around pung the individual plan intoacon and idenfying potenal soluons to any barriers. Describing and invesgangdifficult night me behaviours specific to the individual with demena.5. Taking care of your own (carer’s) sleep: Using the informaon collected on difficult night-me behaviours to create strategies for managing these difficules. Carer managing theirown sleep, including ways to challenge unhelpful thoughts and feelings and make me forthemselves.6. What works? Strategies for the future: This focuses upon what carers have found useful.From this an individualised sleep acon plan is finalised, which includes strategies for boththe person with demena and the carer.Each session, aer the first one, followed a similar structure. They began with a recap of lastweek’s session, and a discussion about what the carer had achieved since then. The new topic wasthen introduced, and the carer and therapist generated relevant, individually-tailored plans thatthe carer could implement in the coming week. At the end of the session, the therapist talked thecarer through a new relaxaon exercise. At the end carers kept their manual, light box andrelaxaon recordings post-intervenon.Figure 1. Structure of DREAMS-START of actiwatch data), behavioral activation,relaxation, and coping skills for families (see,Figure 1 for details of DREAMS-START). Trainedand clinically supervised psychology graduatesdelivered it to carers at home, unless the carerrequested to have sessions elsewhere. Each sessionlasted about 1 h, and took place approximatelyweekly, at a time convenient to the carer. Thesessions were delivered to the family carer, unlessa paid carer was with the person around the clock,and therefore would be most likely to implementstrategies. We encouraged the carers to have thesessions by themselves, so they could talk withoutpotentially distressing the person with dementia.However, if people with dementia wanted toparticipate they were included in the sessions.Our clinical psychologist, PR, met each team oftherapists for 1.5 h of group supervision fortnightly.Additionally, she was available for individualsupervision, requested by the psychology graduateson an ad hoc basis, or occasionally initiated bythe investigators. If the therapists had any urgentconcerns they approached PR, CC, or GL (e.g. ifrisk issues arose). The group supervision formatwas to ensure good case management, clinical skillsdevelopment and safe practice, and to maximizepeer support.PR devised a fidelity checklist comprising themost important components of each session.Therapist’s audio-recorded one session selectedat random by the trial manager per participant.The other therapist in the same team then rated from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at254 G. Livingston et al.the recording for “keeping the session to time,”“keeping the carer focused on the manual,”“keeping the carer engaged in the session,” and“managing the concerns of the carer.” Scoresranged from 1 “not at all” to 5 “very focused.”All participants received TAU. This wasexpected to vary between trusts and patients,but be in line with the NICE dementia guidelines(NICE/SCIE, 2006), consisting of assessment,diagnosis, symptomatic interventions, riskassessment and management, and information.These included medication; cognitive stimulationtherapy; neuropsychiatric symptoms treatment;driving advice; medical identification (ID)bracelets; capacity assessment and adviceregarding power of attorney; referral to dementiaadvisors/navigators for signposting and to socialservices for personal care, tableware, day centerand financial advice, and carer support (START insome trusts). We also gathered details of servicesthe JDR volunteers received.After the follow-up data had been collected,the control group participants received a summaryof the baseline actiwatch data, with advice onimproving sleep.OutcomesTrained research team member, masked to groupassignment, assessed patients for outcomes. Theprimary outcomes were feasibility of recruitmentand treatment adherence. These were assessed by(1) the proportion of eligible carers consented;(2) the proportion randomized after baselineassessment; and (3) acceptability of DREAMSSTART by adherence. The secondary outcomeswere referral rates; follow-up rates at three months;all psychotropic medication prescription (to definerescue medication’s role); reported side effects: comorbid physical illnesses and patient falls; and feasibility and acceptability of interview schedules andactigraphy assessed through instrument completionrates.We collected sleep measures (interview andactigraphy) before randomization and three monthsafter. Actigraphy estimates sleep and wakefulnessfrom movement and time, and has been usedpreviously in a small trial to improve sleep forpeople with dementia (McCurry et al., 2005).We asked carers to define each participant’s sleepanalysis window, either by recording the personwith dementia’s bed and rise times in a sleepdiary, or by pressing the actiwatch’s “event marker”button. The SDI (Tractenberg et al., 2003) is avalid and reliable seven-item measure of a personwith dementia’s sleep disturbance. Scores ≥4 onindividual items equate to clinically significant sleepdisturbance and study eligibility. The SDI total(sum of items) score ranges from 0 to 84. Thealternative SDI mean global score is calculatedas the mean frequency × mean severity (range0–12). Daytime sleepiness was measured usingthe validated eight-item Epworth Sleepiness Scale(ESS) (Johns, 1994; Spira et al., 2012). Possiblescores range 0–24, with >10 indicating excessivesleepiness. We measured overall neuropsychiatricsymptoms using the neuropsychiatric inventory(NPI) (Cummings et al., 1994) at baseline andthree months. This validated instrument has 12neuropsychiatric symptoms domains (scored 0–12). The possible highest score is 144. Higherscores mean increasing severity. Quality of life wasmeasured using the DEMQOL-Proxy (Smith et al.,2007), a 31-item interviewer-administered validand reliable instrument for dementia, which can beused with the Client Service Receipt Inventory tocalculate cost-effectiveness (Mulhern et al., 2012).We collected data on the family carers, includingsleep, commonly disrupted by person with dementia’s sleep–wake patterns, using the validated,reliable Pittsburgh Sleep Quality Index (PSQI)(Buysse et al., 1989), and the Sleep ConditionIndicator (SCI) (Espie et al., 2014), an eightitem scale, characterizing sleep both dimensionallyand against insomnia disorder criteria. Mood wasmeasured using the validated, reliable HospitalAnxiety and Depression Scale (HADS) (Bjellandet al., 2002; Snaith, 2003), and subjective burdenusing the validated Zarit Burden Interview (ZBI)(Zarit et al., 1980). Carer’s quality of life wasmeasured on the Health Status Questionnaire-12(HSQ-12), a 12-item scale (Pettit et al., 2001;Barry et al., 2007).Sample sizeWe estimated that with 40 intervention participants(larger to allow a more precise estimate of proportion adhering to intervention) and 20 controls,we would achieve the following 95% confidenceinterval (CI) for our expected adherence andparticipation estimates:1. Proportion of participants adhering to intervention– expected value 75%, 95% CI: 59–87%2. Proportion of appropriate referrals consenting tothe trial – expected value 50%, 95% CI: 41–59%This sample size was also judged as sufficientfor estimating the standard deviation as requiredfor the sample size calculation in the main trial.We envisaged in our protocol that it wouldbe actigraphy but that we would be open tochange this from information in the trial, and wewould also have information to calculate sample from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available atDREAMS-START to improve sleep in dementia 255size using the scales in the trial. Our estimatedrecruitment referral rate was approximately sixpotential participants per week. Two out ofthese were expected to be suitable and agreeto participate. Our expected follow-up rate wasapproximately 80%.The 95% CI for expected adherence andparticipation estimates would provide acceptableranges to inform continuation to the main trial. Wespecified our “stop–go” measures would be relatedto the proportion adhering, with ≥70% meaningproceed to a main trial.AnalysisFor all eligible patients and carers, we summarizedrecruitment site and sex of the person withdementia, and carer sex and relationship tothe person with dementia comparing those whoconsented with those who did not. We calculatedwith 95% CI: proportion of screened patientswho were eligible for the trial, proportion ofeligible referrals consenting to the trial, proportionof participants in each randomized group whodropped out or were lost to follow-up, proportion ofparticipants in the intervention group who adheredto DREAMS-START (attended a ≥4/6 sessions),median number of sessions attended by those inthe intervention group. The frequency (%) ofparticipants in each randomized group who hadtaken psychotropic drugs and melatonin duringthe three months prior to the baseline and followup was calculated. Frequency (%) of co-morbidphysical illnesses and patient falls was summarizedby randomized group at baseline and three months.We summarized socio-demographic characteristics and baseline and follow-up actigraphy measures and other scores using means (with standarddeviations), medians (with interquartile ranges,IQR), counts and proportions, as appropriate. Theactigraphy data were analyzed using MotionWareSoftware 1.1.25. We defined each sleep period,using the bed and rise times in the sleep diaryor by event markers. We used the carer’s verbalreport when neither was available. When unsure,two researchers edited a sleep period and reacheda consensus. We removed from non-parametriccircadian rhythm analysis (NPCRA) periods ofmissing data >3 h and excluded sleep data basedon fewer than seven nights. For exploratory sleepanalyses, we defined “core night-time” as midnightto 6 am.The number of participants with available datawas summarized for each outcome.Follow-up carer and patient questionnaire scoreswere compared between randomized groups usingregression models to provide estimates of theeffect of DREAMS-START with 95% CI (e.g.difference in means), adjusted for baseline scoreand site. Ordinary least squares regression wasused except for SDI scores where assumptions ofthe model were violated and quantile (median)regression methods were used. Formal analysisfor sleep measures at 3 months was focused onsleep efficiency (time asleeptime in bed), relativeamplitude, activity count for most restful hours andactivity count for most active hours (predefined inanalysis plan). Regression models were adjusted forbaseline score and site. Ordinary least squares wereused where appropriate, otherwise estimates wereobtained from quantile (median) regression.ResultsParticipant recruitment and flowThe flow of participants through the trial isdescribed using a CONSORT diagram (Figure 2).We were referred 123 people through memoryclinics from 04/08/2016 to 03/04/2017. We stoppedrecruiting when we were confident that we wouldhave sufficient randomized participants. The studywas open to JDR recruitment from 22/11/2016to 24/03/2017. We found 140 potential people inJDR, who lived in the area and had registeredas having dementia and a family supporter, andcontacted 27 using their preferred means (25 byemail; 2 by phone) from 05/12/2016 to 21/03/2017.We followed up email by phone calls; 19 did notrespond, 3 were ineligible (2 had no sleep problems;1 was moving out of London), for one person wecould only speak to the person with dementia andwere unable to determine eligibility. 95/120 (79%;95% CI 71–86%) people assessed were eligible.Comparison of those who consented withthose who did notTable 1 compares consenters and non-consentersdemographic details and shows the study had goodexternal validity. The care recipients who did notconsent were, however, more often male.Table 2 compares demographic and diagnosticdetails of the randomized groups. Overall there wasa good demographic mix with recruitment from arange of ethnicities, age groups and relationshipbetween the person with dementia and their carer.There was a range of diagnoses with most peoplehaving AD or a mixed dementia. Most primarycarers (45; 73%) lived with the person withdementia, 4 lived with another family member,so 49/62 (79%) lived with family members, 6(10%) had paid carers living with them, and 7(11%) had family carers but lived alone. The carers from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at256 G. Livingston et al.Referrals from memory clinics (n=123)Approached from JDR (n=27)Assessed for eligibility (n=120)Not assessed for eligibility (n=30)[died (n=1), unable to talk to carer (n=5),uncontactable (n=24)]Excluded (n= 58)Not meeting inclusion criteria (n=25)Declined to participate (n=32)Other reasons (n=1)[Withdrew before randomisation (n=1)]Lost to follow-up (n=4)Withdrawn (n=3)Died (n=1)Discontinued intervention 3 h of missing data). Only 1/62 participantswith dementia randomized did not wear the watchfor ≥7 nights at baseline. The sleep diary or eventmarkers were used by 50/62 (81%) of randomizedparticipants to record the person with dementia’sbed and rise times. Of the remaining 12, 8 (13%)gave a verbal report of sleep pattern, while 4 (6%)did not. Table 6 shows the follow-up sleep data.Overall, there was follow-up data for 49/62 (79%)randomized; 6/57 (10.5%) randomized and still inthe study refused to wear the watch again; 49/51(96%) of those who did wear it had actigraphy datafor at least seven days. Carers of 42 (82%) of these from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available atDREAMS-START to improve sleep in dementia 259Table 3. Summary of comorbid illness and sideeffects by randomized groupintervention tau…………………………………………………………………………………………………………………………………….. BaselineFallsGastroNeurologicalInfectionsOther3 MonthsFallsGastroNeurologicalInfectionsOther(N = 42)17 (40%)21 (50%)19 (45%)14 (34%)6 (14%)(N = 38)16 (42%)16 (42%)15 (39%)18 (47%)9 (24%)(N = 20)7 (35%)9 (45%)10 (50%)6 (30%)2 (11%)(N = 18)7 (39%)4 (22%)10 (56%)5 (28%)9 (50%) Notes: The numbers are frequency (%).participants provided their relative’s bed and risetimes via the sleep diary or event markers. Of theremaining 11, eight (16%) provided a verbal reportof their relative’s sleep pattern.DiscussionThis is the first non-pharmacological randomizedcontrolled trial for sleep problems in dementiapowered to find if the intervention was feasibleand acceptable. It met predetermined criteria toproceed to a full trial. It was not powered forefficacy and our efficacy results are indicative ratherthan definitive. One measure of acceptability is theadherence rates to the intervention, and a goodadherence rate is a matter of judgment, with clinicaltrials of medication for people being treated forchronic illnesses reporting rates of 43–78% (Osterberg and Blaschke, 2005). Our predeterminedlevel in this trial was a relatively high rate of75%, which was exceeded. In addition, most ofthose in the intervention group found it acceptablein that they attended all sessions once they hadstarted.This trial shows that it is possible to deliver acomplex manual-based intervention and measuresfidelity. These results indicate that it shouldbe possible to deliver the intervention reliablyin future. High fidelity was achieved measuredthrough predetermined fidelity check lists andsessions were recorded and fidelity measured.While each possible intervention was mentioned,therapists and carers were explicitly told that ifsomething was not applicable then to mention itas a principle but not go into detail or make it agoal e.g. if someone was wheelchair bound not tosuggest they went out walking daily. Recruitmentrates and consent were high from those referred byhealth professionals but we found that a researchregister may be a less fruitful source of recruitmentpossibly because people may register and then theircircumstances may change. Generally, 80% followup or above is regarded as satisfactory, and weachieved 92%. JDR did not meet our criteria forrecruitment.At follow-up, less than 80% of people randomized had enough actigraphy data to analyze andonly 68% had sleep diary/event marker data. Weoriginally envisaged that actigraphy data wouldbe the full trial primary outcome, but we areunlikely to have the level of data required forthis. Families thought the person they looked afterwas sleeping better, but the actigraphy resultsdid not support this. Actigraphy infers “sleep”from lack of activity during the sleep window.Movement is in contrast directly measured. Thereis little actigraphy validation data in people withdementia, who may frequently stay still whileawake or may sleep during the day. It has beenvalidated against polysomnography (PSG), withgood accuracy at detecting sleep (96.5%), butnot at detecting wakefulness (32.9% accuracy)throughout the age range. Validity worsened withincreasing age (Marino et al., 2013). In olderwomen without dementia (mean age 69 years)actigraphy results were unacceptable for thosewith low sleep efficiency (Taibi et al., 2013). Thealgorithms are also not designed to detect daytimesleepiness. This suggests that it is important thatbetter methods of measuring sleep in people withdementia are measured and it is possible thatWiFi enabled headbands may work better. Therewas, however, indications from actigraphy, thatintervention participants were more active duringthe day and less active at night. These accord withthe validated instruments results.We achieved high completion rates of thequestionnaire measures using carers as informants.The scores on the instruments are useful toinform the design of a full trial. The SDI wascompleted at follow-up by 90% of those initiallyrecruited to the trial. This appeared to be the mostpractical way to measure sleep for future studiesin this area. Summary data for the carer-reportedinstruments indicated generally better scores forthe intervention group, including a significantimprovement in both quality of life of peoplewith dementia and daytime sleepiness despite thesmall numbers and that the comparator group,TAU were receiving secondary care interventions.There was also no increase in numbers usingrescue medication or indication of important harmsin terms of side-effects in either group. Theconsistency in the direction of all the results from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at260 G. Livingston et al.Table 4. Completion rates and scores of baseline and three month follow-up patient and carer validatedinterview measures by randomized groupperson with dementia baselineintervention( N = 42) tau ( N = 20)……………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………….SDI global score Median (IQR) 2.57 (1.35 to3.67)2.98 (1.85 to3.96)SDI total score Median (IQR) 26.5 (19 to 36) 32.5 (20.5 to 40)ESS total score Mean (SD) 9.95 (6.02) 8.85 (6.31)NPI total score Mean (SD) 42.02 (23.17) 46.90 (23.48)(N = 41)DEMQOL-Proxy Mean (SD) 87.57 (10.73) 88.51 (10.14)carer baseline intervention(N = 42)tau (N = 20)PSQI global score Mean (SD) 9.22 (4.08)(N = 41)10.40 (4.52)SCI total score Mean (SD) 15.32 (8.22)(N = 41)13.50 (5.92)HADS scores Anxiety: Mean(SD)8.17 (4.66) 9.30 (3.80)Depression: Mean(SD)5.24 (4.33) 7.65 (4.60)Total: Mean (SD) 13.40 (8.35) 17.05 (7.80)ZBI score Mean (SD) 37.69 (18.39) 38.30 (19.27)HSQ scores Physical health:Mean (SD)67.89 (32.79)(N = 41)52.50 (39.47)Mental health:Mean (SD)60.95 (23.17) 52.33 (24.02)person with dementia three month intervention(N = 42)tau (N = 20) adjusted tr effect(i-tau) (95% ci)bSDI global score Median (IQR) 0.92 (0.49 to2.94)2.43 (0.82 to3.88)-0.30 (-1.42 to 0.82)aSDI total score 0to 84Median (IQR) 16 (9 to 29)(N = 38)30 (14 to 37)(N = 18)-7 (-17.53 to 3.53)aESS total score Mean (SD) 7.17 (5.87)(N = 36)9.00 (7.55)(N = 18)-2.86 (-5.54 to -0.17)(N = 54)DEMQOL-Proxy Mean (SD) 93.52 (10.12)(N = 37)87.07 (10.22)(N = 18)7.08 (2.25 to 11.91)(N = 55)NPI total score Mean (SD) 38.69 (23.57)(N = 36)44.72 (23.22)(N = 18)-1.99 (-11.66 to 7.68)(N = 54)carer three months intervention(N = 42)tau (N = 20) adjusted tr effect(i-tau) (95% ci)bPSQI global score Mean (SD) 9.37 (4.16)(N = 38)9.5 (4.49)(N = 18)1.03 (-1.05 to 3.11)(N = 55)SCI total score Mean (SD) 15.45 (7.45)(N = 38)14.53 (8.54)(N = 19)-0.41 (-3.75 to 2.93)(N = 56)HADS scores Anxiety Mean (SD) 8.76 (5.57) 9.05 (4.22) 1.13 (-0.31 to 2.56)Depression Mean (SD) 5.71 (4.43) 8.79 (4.88) -1.05 (-3.01 to 0.91)Total Mean (SD) 14.47 (9.20)(N = 38)17.84 (8.43)(N = 19)0.51 (-2.39 to 3.42)(N = 57)ZBI score Mean (SD) 36.5 (17.07)(N = 38)42.16 (16.45)(N = 19)-5.32 (-9.83 to -0.82)(N = 57)HSQ scores PhysicalhealthMean (SD) 68.42 (32.60)(N = 38)54.39 (35.94)(N = 19)3.12 (-12.27 to 18.52)(N = 57)MentalhealthMean (SD) 55.96 (26.19)(N = 38)48.07 (21.15)(N = 19)1.25 (-9.52 to 12.02)(N = 57)IQR = interquartile range; SD = standard deviation.aQuantile (median) regression.bEstimates are from models adjusted for baseline score and site. Regression is OLS unless otherwise indicated. from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available atDREAMS-START to improve sleep in dementia 261Table 5. Baseline sleep and non-parametric circadian rhythm analysis measures by randomized group interventionN = 41tauN = 20sleep measures ………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………. Sleep efficiency (%)aAverage sleep time (min)Average wake time (min)Average time of lightsout/bedtime (24 h)bMedian (IQR)Median (IQR)Median (IQR)Median (IQR)76.90 (66.60 to 83.00)448.00 (376 to 503)111 (77 to 152)22:26 (22:07 to 23:08)80.68 (70.75 to 86.30)442.00 (388 to 568.50)114 (58.5 to 147.5)22:25 (21:37 to 23:25) Average time of fallingasleep (24 h)bMedian (IQR) 22:43 (22:07 to 23:40) 22:45 (22:04 to 00:16)Average time of waking up(24 h)Median (IQR) 08:10 (07:41 to 8:54) 08:20 (07:48 to 8:46)Average time of getting up(24 h)Median (IQR) 08:14 (07:52 to 09:00) 08:23 (07:56 to 08:56)Average time in bed (h) Mean (SD) 9.91 (1.35) 10.01 (1.80)non-parametric circadian rhythm analysis measuresN = 42 N = 20Relative amplitude Median (IQR) 0.71 (0.48 to 0.82) 0.78 (0.56 to 0.90)Inter-daily stability Mean (SD) 0.38 (0.16) 0.43 (0.17)Intra-daily variability Mean (SD) 1.11 (0.40) 1.03 (0.32)L5 – average activity countfor five most restful hoursMedian (IQR) 1381 (572 to 1919) 761 (358.5 to 1616)Start hour of five mostrestful hours (24 h)bMedian (IQR) 01:00 (00:00 to 03:00) 01:00 (00:00 to 02:00)Median (IQR) 8155 (5127 to 12281) 8212.0 (3730.5 to14545.5)Start hour of ten mostactive hours (24 h)Median (IQR) 10:00 (8:00 to 12:00) 8:30 (8:00 to 11:00)core night-time sleep measures N = 41 N = 20 Sleep efficiency (%)aAverage sleep time (min)Average wake time (min)Median (IQR)Median (IQR)Median (IQR) N = 39) 300 (246 to 331) (N = 19)N = 39) 52 (26 to 76) (N = 19)Notes: Sleep and core night-time measures are for those ≥7 nights of data available. Non-parametric circadian rhythm analysis (NPCRA)data are those calculated with all 24-h periods with >3 h of missing data excluded.IQR = interquartile range; SD = standard deviation.aTime asleeptime in bed.bTo calculate summaries, times have been ordered from midday one day until midday the next day.suggests they may be real. It is important if ourintervention reduces daytime sleepiness, in contrastto sedative medication, which sometimes increasesit. Similarly, quality of life is an important outcomefor someone with dementia, as an interventionmay improve one domain while reducing overallquality of life. Correspondingly, the carers in theintervention group indicated reductions in stressand burden.The PPI judged the important outcomes werethat the person with dementia was less restlessat night, more awake during the day, disturbedthe carer less and seemed happier (which theSDI measures). In these circumstances, theywere unsure that actigraphy added information,or was accurate. Although some carers weredisappointed with actigraphy feedback, the carersand therapists found the actigraphy informationfrom baseline helped to delineate the rest-activitypattern and help make a plan. We would continueto incorporate this in the intervention manual in afull trial.Future trials and interventions require a degreeof flexibility about who receives the intervention.We had expected we would deliver an interventionmainly to family carers alone, but the interventionwas delivered to family carers, paid carers, andsometimes to both. Frequently, the person withdementia was also included, and in one case halfof the sessions were delivered to a person withdementia alone. Delivering the intervention withthe person with dementia present, although notalways problematic, presented challenges for thetherapists; for example, when the person withdementia denied having any difficulties with sleep.One person with dementia had some sessions alone,but could not retain the information and make useof the sessions. In any future trial, we suggest thatpeople with dementia could jointly participating inthe intervention sessions, with additional training from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at262 G. Livingston et al.Table 6. Three-month sleep and non-parametric circadian rhythm analysis measures by randomized group s l e e p m e a s u r e sN = 32i n t e rv e n t i o nN = 17tauN = 49e f f e c t (i-tau)(95% c i) a……………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………….Sleep efficiency (%)bMedian (IQR)76.05 (65.0 to 81.9)78.6 (72.1 to 82.3)0.39 (-4.89 to5.68)c a d j u st e d t rAverage sleep time(min)Median (IQR) 418.5 (383.5 to 506) 475 (396 to 534)Average wake time(min)Median (IQR) 112.5 (94.5 to157.5)105 (88 to 133)Average time of lightsout/bedtime (24h)dMedian (IQR) 22:15 (21:36 to22:50)22:03 (21:21 to23:28)Average time offalling asleep(24 h)dMedian (IQR) 22:49 (21:59 to23:23)22:22 (21:39 to23:55)Average time ofwaking up (24 h)Median (IQR) 08:04 (07:16 to08:38)08:09 (07:38 to08:41)Average time ofgetting up (24 h)Median (IQR) 08:06 (07:20 to08:44)08:17 (07:48 to08:43)Average time in bed(h)Mean (SD) 9.89 (1.83) 10.05 (1.83)non-parametric circadian rhythm analysis measuresN = 34 N = 17 N = 51Relative amplitude Median (IQR) 0.71 (0.49 to 0.87) 0.76 (0.51 to 0.90) -0.02 (-0.10 to0.06)cInter-daily stability Mean (SD) 0.38 (0.18) 0.47 (0.19)Intra-daily variability Mean (SD) 1.04 (0.38) 0.98 (0.40)L5 – average activitycount for five mostrestful hoursMedian (IQR) 1067 (621 to 2003) 981 (493 to 1940) 76.88 (-521.40 to675.16)cStart hour of fivemost restful hours(24h)dMedian (IQR) 01:00 (00:00 to03:00)01:00 (01:00 to02:00)M10 – averageactivity count for10 most activehoursMedian (IQR) 8247 (4258 to12155)8132 (5960 to17158)-198.29 (-1717.83to 1321.25)cStart hour of tenmost active hours(24 h)Median (IQR) 10:00 (8:00 to11:00)9:00 (8:00 to 10:00)core night-time sleep measures N = 32 N = 17Sleep efficiency (%)b Median (IQR) 79.0 (63.75 to 82.6) 80.9 (69.8 to 86.5)Average sleep time(min)Median (IQR) 285 (243 to 303)(N = 30)291 (251 to 311)(N = 17)Average wake time(min)Median (IQR) 62 (43 to 80)(N = 30)49 (39 to 70)(N = 17)Notes: Sleep and core night-time measures are only for those with at least seven nights of data available. NPCRA (non-parametric circadianrhythm analysis) data are those calculated with all 24-h periods with >3 h of missing data excluded.IQR = interquartile range; SD = standard deviation.aEstimates are from models adjusted for baseline score and site. Regression is OLS unless otherwise indicated.bTime asleeptime in bed.cQuantile (median) regression.dTo calculate summaries, times have been ordered as night time i.e. from midday to midday the following day.for the therapists on delivering sessions withpeople with dementia present and managing anyresulting conflict and interpersonal challenges.However, we recommend, only delivering theintervention to people with dementia with a careralso participating.Strengths and limitationsWe had envisaged that someone with sleepdisturbance and dementia would have a paid orfamily member with them at night to ensure safety.This was not always the case. When people livedalone, the carers (whether family or paid) were from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available atDREAMS-START to improve sleep in dementia 263unable to implement strategies, such as a scheduledbedtime or wind down routine. It was also difficultto gain reliable information about the sleep patternsof people with dementia living alone. We wouldtherefore suggest excluding people without a nighttime carer in a full trial.When paid carers attended the intervention theywere also able to implement strategies. Workingwith them may be important to allow people toremain living at home – care agencies require fullday rates and two carers, if the carer is disturbedfrequently during the night. Since it appearedfeasible and is potentially useful, we would includepaid carers if people with dementia and theirfamilies so wished in a full trial.The trial recruited people from London only,limiting external validity. Although male carerecipients were more likely to refuse, we succeededin recruiting carers from either gender, a rangeof age groups, types and severity of dementia,relationships to the care recipient, marital status,and educational backgrounds. In particular, werecruited about 35% of people of minority ethnicstatus despite most studies finding they are underrepresented. A high proportion of carers remainedin the study. Although the outcome assessors weremasked to outcome, the participants were not. Itis possible that some carers might have felt thatthey had to report a positive result to please theresearcher interviewing them, who was not thetherapist, but in most cases had met them forscreening and baseline assessment. However, theinstruments are validated, and in other studiescarers have frequently reported no beneficial effecton these measures (Livingston et al., 2017). If theperson with dementia lived alone, the carer wouldhave been likely to report their relative’s memoryof their sleep and this assessment may be lessreliable.This study finds that DREAMS-START isa feasible and acceptable intervention. Highfidelity to the intervention suggests it can bereliably delivered by trained, supervised psychologygraduates. The evidence from the validatedquestionnaires suggests there is potential forefficacy in improving sleep disturbance and qualityof life of the person with dementia, and forreducing family carers’ stress. We are able to specifyprimary and secondary outcomes for an efficacytrial and calculate power. This augurs well for a fulltrial.Conflict of interestGL is on the NIHR post-doctoral fellowshipboard. GL and PR have grants from NIHR/ESRC,NIHR HTA, and Alzheimer’s Society and aresupported by North Thames NIHR CLAHRC.CC and GL are supported by UCL NIHRBRU. SK reports grant funding from NIHRHTA, NIHR Oxford Biomedical Research Centre,Arthritis Research UK, and the Dr Mortimer &Theresa Sackler Foundation. JP is an employeeof Alzheimer’s Society. CE reports grant fundingfrom NIHR HTA, NIHR Oxford BiomedicalResearch Centre, the Wellcome Trust, and DrMortimer & Theresa Sackler Foundation. CEreports grants from NIHR HTA during the conductof the study, grants from NIHR HTA, grantsfrom NIHR Oxford Biomedical Research Centre,grants from The Wellcome Trust, grants fromEducation Endowment Foundation, grants fromDr Mortimer & Theresa Sackler Foundation,outside the submitted work. In addition, CEhas a patent U.S. Patent Application SerialNo. 14/172,347 INTERACTIVE SYSTEM FORSLEEP IMPROVEMENT: Docket No. HAME-001 issued and CE is co-founder and Chief MedicalOfficer of Big Health Ltd, which has developedSleepio, an online digital CBT intervention forinsomnia. CE is a shareholder in Big HealthLtd. The other authors declare that they have nocompeting interests.Description of author’s rolesGill Livingston wrote the application and protocoland was Chief Investigator. Gill Livingston,Penny Rapaport, Claudia Cooper, Julie Barber,Simon Kyle, Colin Espie, and Rossana Horsleyconceived and designed the study. Penny Rapaport,Simon Kyle, Gill Livingston, and Claudia Cooperdesigned and revised the manual. James Pickett coordinated the PPI contribution, which was led byRossana Horsley. Kirsi Kinnunen, Lucy Webster,and Brendan Hallam screened the participantsfor eligibility and consented them and gatheredbaseline and follow-up quantitative data. They alsoanalyzed the actigraphy data. Brendan Hallam andLucy Webster delivered the intervention. PennyRapaport supervised the intervention delivery.Julie Barber analyzed the quantitative data. KirsiKinnunen was the trial manager and drafted theCONSORT diagram. Gill Livingston wrote thefirst draft of the paper; all authors revised the reportcritically for important intellectual content and gavefinal approval of the version to be published.Gill Livingston will act as guarantor. She hadfull access to all the data in the study and hadfinal responsibility for the decision to submit forpublication.There was no role of a medical writer in thepaper. from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at264 G. Livingston et al.AcknowledgmentsWe would like to thank the participants andreferrers. We would also like to thank the membersof the Trial Steering Committee – Esme MonizCook (Chair), Kate Maxmin, Judy Leibowitz,Sue Boex and John Cape, and Adam Kadrifor delivering the intervention. The DREAMSSTART research team acknowledges the supportof the National Institute for Health Researchthrough the North Thames Clinical ResearchNetwork. This project was funded by the NationalInstitute for Health Research Health TechnologyAssessment Programme (14/220/06). Camden andIslington NHS Foundation Trust and Barnet,Enfield and Haringey Mental Health NHS Trustfunded Excess Treatment Costs. The study sponsorhad no role in the study design; the collection,analysis, and interpretation of data; the writing ofthe report; or in the decision to submit the paperfor publication.Supplementary materialTo view supplementary material for thisarticle, please visit, T. L., Kaiser, K. L. and Atwood, J. R. (2007).Reliability, validity, and scoring of the Health StatusQuestionnaire-12 version 2.0. Journal of NursingMeasurement, 15, 24–35.Bjelland, I., Dahl, A. A., Haug, T. T. and Neckelmann,D. (2002). The validity of the hospital anxiety anddepression scale. an updated literature review. Journal ofPsychosomatic Research, 52, 69–77.Buysse, D. J., Reynolds, C. F. III, Monk, T. H., Berman,S. R. and Kupfer, D. J. (1989). The Pittsburgh sleepquality index: a new instrument for psychiatric practice andresearch. Psychiatry Research, 28, 193–213.Cummings, J. L., Mega, M., Gray, K.,Rosenberg-Thompson, S., Carusi, D. A. andGornbein, J. (1994). The neuropsychiatric inventory:comprehensive assessment of psychopathology in dementia.Neurology, 44, 2308–14.Dauvilliers, Y. (2007). Insomnia in patients withneurodegenerative conditions. Sleep Medicine, 8, S27–S34.Espie, C. A., Kyle, S. D., Hames, P., Gardani, M.,Fleming, L. and Cape, J. (2014). The sleep conditionindicator: a clinical screening tool to evaluate insomniadisorder. BMJ Open, 4, e004183.Forbes, D., Blake, C., Thiessen, E., Peacock, S. andHawranik, P. (2014). Light therapy for improvingcognition, activities of daily living, sleep, challengingbehaviour, and psychiatric disturbances in dementia.Cochrane Database of Systematic Reviews, CD003946. DOI:10.1002/14651858.CD003946.pub4.Hughes, C. P., Berg, L., Danziger, W. L., Coben, L. A.and Martin, R. L. (1982). A new clinical scale for thestaging of dementia. The British Journal of Psychiatry, 140,566–72.Johns, M. W. (1994). Sleepiness in different situationsmeasured by the Epworth sleepiness scale. Sleep, 17,703–10.Ju, Y., Lucey, B. P. and Holtzman, D. M. (2014). Sleepand Alzheimer disease pathology – a bidirectionalrelationship. Nature Reviews Neurology, 10, 115–9.Kinnunen, K. M., Vikhanova, A. and Livingston, G.(2017). The management of sleep disorders in dementia:an update. Current Opinion in Psychiatry, 30, 491–7.doi:10.1097/YCO.0000000000000370Kyle, S. D., Espie, C. A. and Morgan, K. (2010). “…Notjust a minor thing, it is something major, which stops youfrom functioning daily”: quality of life and daytimefunctioning in insomnia. Behavioral Sleep Medicine, 8,123–40.Livingston, G. et al. (2013). Clinical effectiveness of amanual based coping strategy programme (START,STrAtegies for RelaTives) in promoting the mental healthof carers of family members with dementia: pragmaticrandomised controlled trial. British Medical Journal, 347.Livingston, G. et al. (2014). Long-term clinical andcost-effectiveness of psychological intervention for familycarers of people with dementia: a single-blind, randomised,controlled trial. Lancet Psychiatry, 1, 539–48.Livingston, G. et al. (2017). Dementia prevention,intervention, and care. Lancet, 390, 2673–734.Marino, M. et al. (2013). Measuring sleep: accuracy,sensitivity, and specificity of wrist actigraphy compared topolysomnography. Sleep, 36, 1747–55.McCleery, J., Cohen, D. and Sharpley, A. (2014).Pharmacotherapies for sleep disturbances in Alzheimer’sdisease. Cochrane Databaseof Systematic Reviews, 21,CD009178.McCrae, C. S., Dzierzewski, J. M., McNamara, J. P.,Vatthauer, K. E., Roth, A. J. and Rowe, M. A. (2016).Changes in sleep predict changes in affect in oldercaregivers of individuals with alzheimer’s dementia: amultilevel model approach. The Journals of Gerontology.Series B, Psychological Sciences and Social Sciences, 71,458–62.McCurry, S. M. and Ancoli-Israel, S. (2003). Sleepdysfunction in Alzheimer’s disease and other dementias.Current Treatment Options in Neurology, 5, 261–72.McCurry, S. M., Gibbons, L. E., Logsdon, R. G.,Vitiello, M. V. and Teri, L. (2005). Nighttime insomniatreatment and education for Alzheimer’s disease: arandomized, controlled trial. Journal of the AmericanGeriatrics Society, 53, 793–802.McCurry, S. M., Gibbons, L. E., Logsdon, R. G.,Vitiello, M. V. and Teri, L. (2009). Insomnia incaregivers of persons with dementia: who is at risk and whatcan be done about it?. Sleep Medicine Clinics, 4, 519–26.Montgomery, P. and Dennis, J. (2003). Cognitivebehavioural interventions for sleep problems in adults aged60+. Cochrane Database of Systematic Reviews, (1),CD003161. from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available atDREAMS-START to improve sleep in dementia 265Moran, M., Lynch, C. A., Walsh, C., Coen, R., Coakley,D. and Lawlor, B. A. (2005). Sleep disturbance in mildto moderate Alzheimer’s disease. Sleep Medicine, 6, 347–52.Morris, J. C. (1997). Clinical dementia rating: a reliable andvalid diagnostic and staging measure for dementia of theAlzheimer type. International Psychogeriatric, 9, 173–6.Mulhern, B. et al. (2012). Improving the measurement ofQALYs in dementia: developing patient- andcarer-reported health state classification systems usingRasch analysis. Value Health, 15, 323–33.(NICE/SCIE) (2006). NIfCEaSCIfE. Dementia: SupportingPeople with Dementia and Their Carers in Health and SocialCare. London: The British Psychological Society andGaskell.Osterberg, L. and Blaschke, T. (2005). Adherence tomedication. The New England Journal of Medicine, 353,487–97.Pettit, T., Livingston, G., Manela, M., Kitchen, G.,Katona, C. and Bowling, A. (2001). Validation andnormative data of health status measures in older people:the Islington study. International Journal of GeriatricPsychiatry, 16, 1061–70.Prince, M., Wimo, A. G. M., Ali, G. C., Wu, Y. T. andPrina, M. (2015). World Alzheimer Report 2015 – TheGlobal Impact of Dementia: An Analysis of Prevalence,Incidence, Cost and Trends. London: Alzheimer’s DiseaseInternational.Sivertsen, B. and Nordhus, I. H. (2007). Management ofinsomnia in older adults. The British Journal of Psychiatry190, 285–6.Smith, S. C. et al. (2007). Development of a new measure ofhealth-related quality of life for people with dementia:DEMQOL. Psychological Medicine, 37, 737–46.Snaith, R. P. (2003). The hospital anxiety and depressionscale. Health and Quality of Life Outcomes, 1, 29.Spira, A. P. et al. (2012). Reliability and validity of thePittsburgh sleep quality index and the Epworth sleepinessscale in older men. The Journal of Gerontology, Series A:Biological Sciences and Medical Sciences, 67, 433–9.Taibi, D. M., Landis, C. A. and Vitiello, M. V. (2013).Concordance of polysomnographic and actigraphicmeasurement of sleep and wake in older women withinsomnia. Journal of Clinical Sleep Medicine, 9, 217–25.Tractenberg, R. E., Singer, C. M., Cummings, J. L. andThal, L. J. (2003). The sleep disorders inventory: aninstrument for studies of sleep disturbance in personswith Alzheimer’s disease. Journal of Sleep Research, 12,331–7.Zarit, S. H., Reever, K. E. and Bach-Peterson, J. (1980).Relatives of the impaired elderly: correlates of feelings ofburden. Gerontologist, 20, 649–55.Zhao, Q. F. (2016). The prevalence of neuropsychiatricsymptoms in Alzheimer’s disease: systematic review andmeta-analysis. Journal of Affective Disorders, 190, 264–71. from University College London (UCL), on 02 Dec 2019 at 18:05:26, subject to the Cambridge Core terms of use, available at


Leave a Reply

Your email address will not be published. Required fields are marked *