Non-pharmacological interventions for agitation in dementia: systematic review of randomised controlled trials

This review article discusses the prevalence of agitation in dementia and the ineffectiveness of medication in treating it. It provides evidence-based strategies for care homes, including person-centred care, communication skills training, and adapted dementia care mapping. The article also suggests future interventions should focus on consistent and long-term implementation through staff training. The study reviewed 33 studies fitting predetermined criteria and assessed their validity and calculated standardised effect sizes (SES).

The numberof peoplewith dementiais risingrapidlywith
increasedlongevity.Althoughdementia’score symptom is
cognitivedeterioration,agitation iscommon,persistentand
distressing. Nearly half of all people with dementia have agitation
symptoms every month,including 30% of those living at home.1
Four-fifths ofthose with clinically significantsymptoms remain
agitated over 6 months,2 and 20% of those initially symptom-free
developsymptomsover 2 years.2 Agitationin dementiais
associated with poorquality oflife,3 because itis unpleasant,
impedes activities and relationships, causes helplessness and anger
in family and paid caregivers,4 and predictsnursinghome
admission,5 wherethe agitated behaviouradversely influences
the environment.4 Severalreviews,includingour previous
systematicreview,6 considered allneuropsychiatricsymptoms’
management together.We found direct behaviouralmanagement
therapies (BMT) with the person with dementia and specific staff
Psychotropic medication was routinely used to treat agitation
but is now discouraged since benzodiazepines and antipsychotics
increasecognitivedecline,10 and antipsychoticscauseexcess
mortality and are oflimited efficacy.11 Similarly,citalopram has
some efficacy but has cardiac side-effects and reduces cognition.12
Cholinesterase inhibitors and memantine appear ineffective.13,14
Preliminary evidence suggests mirtazapine may reduce agitation.15
One RCT (not placebo-controlled)found analgesicsimproved
agitation in people with dementia, with an effect size comparable
to antipsychotics.16
Effective agitation management could in theory improve the
quality of life of people with dementia and their caregivers, reduce
distress,decreaseinappropriatemedication,enablepositive
relationshipsand activities,delayinstitutionalisation and be
cost-effective.We aimed therefore to review systematically the
evidence for non-pharmacologicalinterventions for agitation in
Non-pharmacologicalinterventions for agitation
in dementia: systematic review of randomised
controlled trials
Gill Livingston,Lynsey Kelly,Elanor Lewis-Holmes,Gianluca Baio,Stephen Morris,
Nishma Patel,Rumana Z.Omar,Cornelius Katona and Claudia Cooper
Background
Agitation in dementia is common,persistentand distressing
and can lead to care breakdown.Medication is often
ineffective and harmful.
Aims
To systematically review randomised controlled trialevidence
regarding non-pharmacologicalinterventions.
Method
We reviewed 33 studies fitting predetermined criteria,
assessed their validity and calculated standardised effect
sizes (SES).
Results
Person-centred care,communication skills training and
adapted dementia care mapping decreased symptomatic and
severe agitation in care homes immediately (SES range
0.3–1.8)and for up to 6 months afterwards (SES range
0.2–2.2).Activities and music therapy by protocol(SES range
0.5–0.6)decreased overallagitation and sensory intervention
decreased clinically significantagitation immediately.
Aromatherapy and lighttherapy did notdemonstrate
efficacy.
Conclusions
There are evidence-based strategies for care homes.Future
interventions should focus on consistentand long-term
implementation through stafftraining.Further research is
needed for people living in their own homes.
Declaration of interest
None.
The British Journalof Psychiatry (2014)
205, 436–442.doi: 10.1192/bjp.bp.113.141119
Review article

professionals.We hand-searched included papers’reference lists
and contacted allauthorsaboutother relevantstudies.We
translated eight non-English papers.
Inclusion and exclusion criteria
We included studiesin any languagethat met the following
criteria:
(a) the participants all had dementia, or those with dementia were
analysed separately;
(b) thestudy evaluated non-pharmacologicalinterventionsfor
agitation,defined asinappropriateverbal,vocalor motor
activity not judged by an outside observer to be an outcome
of need,18 encompassing physicaland verbalaggression and
wandering;
(c) agitation was measured quantitatively;
(d) a comparator group was reported or agitation was compared
before and after the intervention.
We excluded studies if every individual was given psychotropic
drugsor someparticipantsreceived medication asthe sole
intervention.In this paper we report the highest-quality studies
– randomisedcontrolledtrials (RCTs) with more than 45
participants – since none of the trials with a smaller sample size
provided a fulland appropriate sample size calculation.
Data extraction
The first20 search results were independently screened by G.L.
and L.K.to assess exclusion procedure reliability.No paper was
excluded incorrectly.All other papers were screened by L.K.and
E.L.H.If exclusion was unclear,L.K., E.L.H.and G.L.discussed
and reached consensus.Data extracted from the papers (by L.K.
and E.L.H.) included methodologicalcharacteristics;description
of the intervention;whether the intervention was applied to the
personwith dementia,family caregiversor staff; statistical
methods;length of follow-up;diagnostic methods;and summary
outcomedata (immediateand longer-term).Paper quality,
including bias,wasscored independently by L.K.and E.L.H.,
discussing discrepancies with G.L.and/or G.B.They used Centre
for Evidence-Based Medicine(CEBM) RCT evaluation criteria
(http://www.cebm.net/index.aspx?o = 1025);this approach gives
points for randomisation and its adequacy,participant and rater
masking,outcomemeasuresvalidity and reliability,power
calculationsand achievement,follow-up adequacy,accounting
for participants,and whetheranalyseswere intention to treat
and appropriate.Possiblescoresrange from 0 to 14 (highest
in person-centred care orcommunication skills(interventions
focused on improvingcommunication with theperson with
dementia and finding out what they wanted),with and without
supervision;dementiacare mapping;aromatherapy;training
family caregiversin behaviouralmanagementtherapiesor
cognitive–behaviouraltherapy (CBT); exercise;cognitive
stimulation therapy;and simulated presence therapy.
Agitation level
We separated studiesaccordingto the inclusion criteriaof
participantsin termsof levelof symptomsof agitation:1, no
agitation symptom necessary forinclusion;2, someagitation
symptomsnecessaryfor inclusion;3, clinicallysignificant
agitation level;4,level unspecified.We used the usual thresholds:
a score above 39 on the Cohen-Mansfield Agitation Inventory
(CMAI),20and a score above 4 on the Neuropsychiatric Inventory
(NPI) agitation scale,1 to denote significant agitation.
Statistical analysis
We decided a priorito meta-analyse when there were three or
more RCTs investigating sufficiently homogeneous interventions
using the same outcome measure,but no intervention met these
criteria.To facilitatecomparisonacrossinterventionsand
outcomes, where possible, we estimatedinterventions’
standardised effectsizes (SES) with 95% confidence intervals.21
In some studies the outcome was measured and reported at several
time-points during the intervention.We used data from the last
time-point to estimate the SES, since individual patient data were
not available to incorporate repeated measures in the calculation.
We also recalculated resultsfor studiesnot directly comparing
intervention and control groups but reporting only within-group
comparisons and with one-tailed significance tests, so some of our
results differ from the originalanalysis.
Results
We found 1916 records,including 33 relevant RCTs with at least
45 participants(Fig. 1). Online TablesDS1 and DS2 list
methodologicalcharacteristics,SES and qualityratings;Table
DS1 containsthe findingsfrom interventionsfor which there
appeared to be adequate evidence,and Table DS2 contains those
for which there was not adequate evidence (either evidence that
they were noteffective orwhere there wassimply insufficient
evidence).
Non-pharmacologicalstudies of agitation in dementia

Livingston et al
activitiesfurtherreducesagitation.Thereis no evidencefor
activities in severe agitation or outside care homes.
Music therapy.Three RCTs,allin care homes,evaluated music
therapy by trained therapists using a specific protocol– typically
involving warming up with a well-known song,listening to and
then joining in with the music.28–30The largeststudy,which
included participants irrespective of agitation level,found music
therapy twice a week for 6 weeks was effective compared with
the usualcare group.28 A second study found a significant effect
in comparison with a reading group,30 and the third found a
efficaciousthan ordinary massage orusualtreatment.Sensory
interventions significantly improved symptomatic agitation and
clinicallysignificantagitationduring the intervention,but
therapeutictouch did notdemonstrateadded advantage,and
there is insufficient evidence about long-term effects or in settings
outside care homes.
Working through care-home staff
Person-centred care, communication skills training and dementia
caremappingall seekto changethe caregiver’sperspective,
102 additionalrecords
found from
other sources
Initialscreening
1916 records
284 fullpapers reviewed
160 papers included
and scored for quality
including bias
33 RCTs with at least
45 participants:
CEBM score 1b:8
CEBM score 2b: 25
133 records
identified from
second search
1632 abstracts excluded:
Not only dementia: 681
Not primary research:360
Not intervention study:318
Not psychological, or behaviouralor sensory
or environmentalintervention:236
Protocolonly: 16
No comparator group:9
No agitation or behaviouraloutcome:11
Conference paper only: 1
124 excluded
No agitation or behavouraloutcome:35
Not only dementia: 23
Not primary research:15
No comparator:12
No quantitative outcome: 17
Conference paper only: 8
Not psychological, behavioural, sensory
or environmentalintervention:3
Multidisciplinary team approach including drug prescription:4
Not intervention study:3
Participants under 50: 2
No outcome:1
Protocolonly: 1
127 not considered in this review
CEBM score 2c or less:107
CEBM score 2b or more: 545 participants:19
CEBM score 2b 445 participants not RCT: 1
1681 unique records
identified from
originalsearch
6
7
6
7
6
7
6
Fig. 1 Study search profile (CEBM, Centre for Evidence-Based Medicine; RCT, randomised controlled trial).

Non-pharmacologicalstudies of agitation in dementia
Dementia caremapping.One large,high-quality carehome
study evaluated dementia care mapping. The researchers observed
and assessed each resident’sbehaviour,factorsimproving well-
being and potentialtriggers;explained the results to caregivers,
and supported proposed change implementation. Severe agitation
decreased during the intervention and 4 months afterwards.35
Effectsizes. Training paid care-home staffin communication
skills, person-centredcare or dementiacare mappingwith
supervision during implementation was significantly effective for
symptomaticand severeagitation immediately (SES = 0.3–1.8)
and for up to 6 months (SES = 0.2–2.2).There was no evidence
in other settings.
Interventions without evidence of efficacy
Working with the person with dementia
Light therapy. Light therapy hypothetically reducesagitation
through manipulating circadian rhythms,typically by 30–60 min
daily brightlightexposure.We included three RCTs,all in care
homes.40–42 Among participantswith some or significant
agitation,light therapyeitherincreased agitation ordid not
improveit. The SES was0.2 (for improvement)to 4.0 (for
worsening symptoms) compared with the controlgroup.There
is therefore no evidence that light therapy reduces symptomatic
or severe agitation in care homes and it may worsen it.
Aromatherapy.The two RCTs of aromatherapy both took place
in care homes.43,44
One large, high-quality blinded study found no
immediate or long-term improvement relative to the control group
for participants with severe agitation.44 The other,non-blinded,
study found significant improvement compared with the control
group.43 When assessorsare maskedto the intervention,
aromatherapy hasnot been shown to reduce agitation in care
homes.
Training family caregiversin BMT. Two high-quality studies
found no immediate or longer-term effect (at 3 months, 6 months
or 12 months) ofeither four or eleven sessions training family
caregivers in BMT for severe or symptomatic agitation in people
with dementia living athome.45,46Two studiestraining family
caregivers in CBT for people with severe agitation also found no
improvementcompared with controls.47,48There isthushigh-
quality evidence thatteaching family caregivers BMT or CBT is
ineffective for severe agitation,but insufficient evidence to draw
conclusions regarding symptomatic agitation.
when the participant was agitated – was not effective.51One study
testing a mixed psychosocial intervention,including massage and
promoting residents’activities ofdaily living skills,did not find
agitationimprovedsignificantlycomparedwith the control
group.52
Standardised effect sizes
Figure 2 illustrates the effect of person-centred care, communication,
dementia care mapping,music therapy and activities in reducing
agitation. Long-term effects (in months) of changing the way care-
givers interact with residents are at least as good as the short-term
effects.35,38
Discussion
This is the first up-to-date systematic review to focus on agitation.
It uniquelyanalyseswhetherthe intervention waspotentially
preventive,by reducingmeanlevelsof agitationsymptoms
including those not clinically significant at baseline or managed
clinically significant agitation;whether effects were observed only
while the intervention wasin place orlasted longer;and the
settings in which the intervention had been shown to be effective:
the community or in care homes.
Effective interventions
Effective interventions seem to work through care staff,partic-
ularly in the long term.Thereis convincingevidencethat
when implementation issupervised,interventionsthataim to
communicatewith peoplewith dementia,helpingstaff to
understand and fulfil their wishes, reduce symptomatic and severe
agitation during the intervention and for 3–6 months afterwards.
This suggeststhattraining paid caregiversin communication,
person-centred care skills or dementia care mapping are clinically
importantinterventions,as shown by a 30% decreasein
agitation43 or a standardised effect size of 0.2,which is clinically
small,0.5 medium and 0.8 large.53
Sensory interventions significantly improved agitation ofall
severitieswhile in place.Therapeutictouch had no added
advantage.We also found replicated,good-quality evidence that
activitiesand musictherapyby protocolreduceoveralland
symptomatic agitation in care homeswhile in place.Although
we were surprised thatindividualised activitieswere no more
effective than prescribed activities, the low numbers in the activity
intervention groups may suggest that it was only those who were
particularlysuited to theactivitywho participated.Thereis
no evidencefor severeagitation.Theory-basedactivities

Livingston et al
databases,translated non-English publications,reduced publicationinsufficientevidence to draw conclusions.Severalinterventions
Woods et al(2009)33
Restlessness (v. control)
Woods et al(2009)33
Restlessness (v. placebo)
Gormley et al(2001)45
Ancoli-Israelet al(2003)40
Ancoli-Israelet al(2003)40
Dowling et al(2007)42
p.m.Light
Dowling et al(2007)42
a.m.Light
Burns et al(2009)41
Burns et al(2009)41
Chenoweth et al(2009)35
Chenoweth et al(2009)35
Chenoweth et al(2009)35
Chenoweth et al(2009)35
Deudon et al(2009)36
Deudon et al(2009)36
McCallion et al(1999)37
McCallion et al(1999)37
McCallion et al(1999)38 Physicalaggression Person-centred care and communication skills
McCallion et al(1999)38 Physicalaggression
McCallion et al(1999)38 Verbalaggression
McCallion et al(1999)38 Verbalaggression
McCallion et al(1999)38 Physicalnon-aggression
McCallion et al(1999)38 Physicalnon-aggression
Lin et al(2011)28
Sung et al(2012)39
Kolanowskiet al(2011)24
Matched to interests
Kolanowskiet al(2011)24
Matched to functionla level
Kolanowskiet al(2011)24
Matched to both
Music therapy with a specific protocol
Activities
Therapeutic touch
Training family caregivers in behaviouralmanagement
for people with dementia living at home
Light therapy
Dementia care mapping
Long-term effect
Short-term effect
78 76 74 72 0 2
Interventions worsen agitation
4
Interventions improve agitation
Fig. 2 Standardised effect size and 95% confidence intervals where calculable of randomised controlled trials compared with controls
for each reported outcome immediately and in the longer term.

Non-pharmacologicalstudies of agitation in dementia
examine which symptoms contributed to this effect and if it were
mood rather than agitation.8
Study implications
Although agitation in dementia has been regarded as due to brain
changes,our findings suggestagitation also arises from lack of
understanding or unmet needs in someone whose dementia makes
them unable to explain or understand this. This is in line with the
need-driven,dementia-compromised behaviour theory of Algase
et al,54 and the hypothesis of Kitwood & Bredin that behaviours
arise from need and occur when care is task-driven not person-
centred(relevantto all neuropsychiatricsymptoms).55 Our
findingssuggestcliniciansshould stop considering agitation as
an entity but instead often as a symptom of lack of understanding
or unmet need that the person with dementia is unable to explain
or understand.This may bephysicaldiscomfortor need for
stimulation,emotionalcomfort or communication.
Future research
More evidence is required aboutimplementing group activities
in care homesoverlongerperiodsto preventagitation.We
recommend the development and evaluation ofa manual-based
training forstaffin care homesemploying interventionswith
evidence forefficacy,to allow translation to differentsettings.
We suggest these interventions should focus on changing culture
to implement programmes permanently.In generalit seems that
there is no evidence about settings outside care homes.The lack
of effective interventions, despite 70–80% of people with dementia
living athome and the potentialof interventions to delay care
homeadmission,suggestsfurtherresearch should startfrom
qualitative interviewsconsidering how agitation isexperienced
by people with dementia living at home,and how their families
manage.This, togetherwith synthesised evidencefrom other
settings,could help in the developmentof a pilotintervention.
Our review may suggest that it should have elements of sensory
stimulation (including music),activities and teaching the family
caregiver communication skills,to change themselves rather than
the person with dementia.
Gill Livingston, MD, Lynsey Kelly, BSc, Elanor Lewis-Holmes, BSc, Unit of Mental
Health Sciences,Gianluca Baio, PhD,Departments ofStatisticalScience and
PRIMENT ClinicalTrials Unit, Stephen Morris, PhD, Nishma Patel, MSc, Department
of Applied Health Research,Rumana Z.Omar, PhD,Departments ofStatistical
Science and PRIMENT ClinicalTrials Unit,Cornelius Katona, MD, Claudia Cooper,
PhD,Unit ofMentalHealth Sciences,University College London,UK
Correspondence: Professor GillLivingston,Charles BellHouse,67–73 Riding
References
1 Ryu SH, Katona C,Rive B,Livingston G.Persistence of and changes in
neuropsychiatric symptoms in Alzheimer disease over 6 months – the LASER-
AD study.Am J Geriatr Psychiatry 2005;13: 976–83.
2 Savva GM, ZaccaiJ, Matthews FE,Davidson JE,McKeith I, Brayne C,et al.
Prevalence,correlates and course of behaviouraland psychological
symptoms of dementia in the population.Br J Psychiatry 2009;194: 212–9.
3 Wetzels RB, Zuidema SU,de Jonghe JFM, Verhey FRJ,Koopmans RTCM.
Determinants of quality of life in nursing home residents with dementia.
Dementia and Geriatric Cognitive Disorders 2010;29: 189–97.
4 Draper B, Snowdon J, Meares S, Turner J, GonskiP, McMinn B, et al.
Case-controlled study of nursing home residents referred for treatment of
vocally disruptive behavior.Int Psychogeriatr 2000;12: 333–44.
5 Morris LW, Morris RG, Britton PG. The relationship between maritalintimacy,
perceived strain and depression in spouse caregivers of dementia sufferers.
Br J Med Psychol 1988;61: 231–6.
6 Livingston G, Johnston K, Katona C, Paton J, Lyketsos CG. Systematic review
of psychologicalapproaches to the management of neuropsychiatric
symptoms of dementia.Am J Psychiatry 2005;162: 1996–2021.
7 National Institute for ClinicalExcellence, SocialCare Institute for Excellence.
Dementia:Supporting People with Dementia and Their Carers in Health and
Social Care. NICE/SCIE,2006.
8 Brodaty H, Arasaratnam C.Meta-analysis of nonpharmacological
interventions for neuropsychiatric symptoms of dementia.Am J Psychiatry
2012;169: 946–53.
9 Kong EH, Evans LK,Guevara JP.Nonpharmacologicalintervention for
agitation in dementia:a systematic review and meta-analysis.Aging Ment
Health 2009;13: 512–20.
10 Bierman EJ, Comijs HC,Gundy CM,Sonnenberg C,Jonker C, Beekman AT.
The effect of chronic benzodiazepine use on cognitive functioning in older
persons:good, bad or indifferent? Int J Geriatr Psychiatry 2007;22:
1194–200.
11 Maher AR, Maglione M,Bagley S,Suttorp M, Hu JH, Ewing B, et al.
Efficacy and comparative effectiveness of atypicalantipsychotic medications
for off-labeluses in adults:a systematic review and meta-analysis.JAMA
2011;306: 1359–69.
12 Porsteinsson AP,Drye LT,Pollock BG,Devanand DP,Frangakis C,IsmailZ,
et al. Effect of citalopram on agitation in Alzheimer disease:the CitAD
randomized clinicaltrial. JAMA 2014; 311: 682–91.
13 Howard RJ, Juszczak E,Ballard CG,Bentham P,Brown RG,Bullock R,et al.
Donepezilfor the treatment of agitation in Alzheimer’s disease. N EnglJ Med
2007;357: 1382–92.
14 Fox C, CrugelM, Maidment I,Auestad BH,Coulton S,Treloar A,et al.
Efficacy of memantine for agitation in Alzheimer’s dementia:a randomised
double-blind placebo controlled trial.PLoS One 2012;7: e35185.
15 Banerjee S,Hellier J, Dewey M, Romeo R,Ballard C,Baldwin R,et al.
Sertraline or mirtazapine for depression in dementia (HTA-SADD):a
randomised, multicentre,double-blind, placebo-controlled trial.Lancet 2011;
378: 403–11.
16 Husebo BS,Ballard C,Sandvik R,Nilsen OB,Aarsland D.Efficacy of treating
pain to reduce behaviouraldisturbances in residents of nursing homes with
dementia:cluster randomised clinicaltrial. BMJ 2011; 343: d4065.
17 Livingston G,Kelly L, Lewis-Holmes E,Baio G, Morris S, Patel N, et al.

Livingston et al
24 Kolanowski A, Litaker M, Buettner L,Moeller J, Costa PT.A randomized
clinicaltrial of theory-based activities for the behavioralsymptoms of
dementia in nursing home residents.J Am Geriatr Soc 2011;59: 1032–41.
25 Kovach C, Cashin S,TaneliY, Dohearty P,Schlidt A, Silva-Smith A.
Effects of the BACE intervention on agitation of demented residents in
longterm care.Gerontologist 2003;43: 233.
26 Cohen-Mansfield J,Jensen B. Do interventions bringing current self-care
practices into greater correspondence with those performed premorbidly
benefit the person with dementia? A pilot study.Am J Alzheimer Dis Other
Demen 2006;21: 312–7.
27 Cohen-Mansfield J, Parpura-GillA, Golander H. Utilization of self-identity roles
for designing interventions for persons with dementia.J Gerontol Ser B:
PsycholSciSoc Sci2006;61B: 202–12.
28 Lin Y, Chu H, Yang CY, Chen CH,Chen SG,Chang HJ,et al. Effectiveness
of group music intervention against agitated behavior in elderly persons
with dementia.Int J Geriatr Psychiatry 2011;26: 670–8.
29 Sung HC, Lee WL,Li TL, Watson R.A group music intervention using
percussion instruments with familiar music to reduce anxiety and agitation
of institutionalized older adults with dementia.Int J Geriatr Psychiatry 2012;
27: 621–7.
30 Cooke ML, Moyle W, Shum DH,Harrison SD,Murfield JE. A randomized
controlled trialexploring the effect of music on agitated behaviours and
anxiety in older people with dementia. Aging Ment Health 2010; 14: 905–16.
31 Hawranik P, Johnston P, Deatrich J.Therapeutic touch and agitation in
individuals with Alzheimer’s disease.West J Nurs Res 2008;30: 417–34.
32 Woods DL, Craven RF,Whitney J.The effect of therapeutic touch on
behavioralsymptoms of persons with dementia.Altern Ther Health Med
2005;11: 66–74.
33 Woods DL, Beck C, Sinha K. The effect of therapeutic touch on behavioral
symptoms and cortisolin persons with dementia.Forsch Komplementmed
2009;16: 181–9.
34 Van Weert JCM, van Dulmen AM,Spreeuwenberg PMM,Ribbe MW,
Bensing JM.Behavioraland mood effects of Snoezelen integrated into
24-hour dementia care.J Am Geriatr Soc 2005;53: 24–33.
35 Chenoweth L, King MT, Jeon YH, Brodaty H, Stein-Parbury J, Norman R, et al.
Caring for Aged Dementia Care Resident Study (CADRES)of person-centred
care, dementia-care mapping,and usualcare in dementia:a cluster-
randomised trial.Lancet Neurol2009;8: 317–25.
36 Deudon A, Maubourguet N, Gervais X, Leone E, Brocker P, Carcaillon L, et al.
Non-pharmacologicalmanagement of behaviouralsymptoms in nursing
homes.Int J Geriatr Psychiatry 2009;24: 1386–95.
37 McCallion P, Toseland RW,Freeman K.An evaluation of a Family Visit
Education Program.J Am Geriatr Soc 1999;47: 203–14.
38 McCallion P, Toseland RW, Lacey D, Banks S. Educating nursing assistants to
communicate more effectively with nursing home residents with dementia.
Gerontologist 1999;39: 546–58.
39 Sloane PD, Hoeffer B,MitchellCM, McKenzie DA,Barrick AL,Rader J, et al.
Effect of person-centered showering and the towelbath on bathing-
associated aggression,agitation,and discomfort in nursing home residents
with dementia:a randomized,controlled trial.J Am Geriatr Soc 2004;52:
1795–804.
40 Ancoli-IsraelS, Martin JL, Gehrman P,Shochat T,Corey-Bloom J,Marler M,
et al. Effect of light on agitation in institutionalized patients with severe
Alzheimer disease.Am J Geriatr Psychiatry 2003;11: 194–203.
42 Dowling GA,Graf CL,Hubbard EM,Luxenberg JS.Light treatment for
neuropsychiatric behaviors in Alzheimer’s Disease. West J Nurs Res 2007; 29:
961–75.
43 Ballard CG,O’Brien JT,Reichelt K,Perry EK.Aromatherapy as a safe and
effective treatment for the management of agitation in severe dementia:
the results of a double-blind,placebo-controlled trialwith Melissa.J Clin
Psychiatry 2002;63: 553–8.
44 Burns A, Perry E, Holmes C,Francis P, Morris J, Howes MJ, et al.
A double-blind placebo-controlled randomized trialof Melissa officinalis
oil and donepezilfor the treatment of agitation in Alzheimer’s disease.
Dementia Geriatr Cogn Disord 2011;31: 158–64.
45 Gormley N,Lyons D,Howard R.Behaviouralmanagement of aggression in
dementia:a randomized controlled trial.Age Ageing 2001;30: 141–5.
46 Teri L, Logsdon RG,Peskind E,Raskind M,Weiner MF,Tractenberg RE,
et al. Treatment of agitation in AD – a randomized,placebo-controlled
clinicaltrial. Neurology 2000;55: 1271–8.
47 Wright LK,Litaker M,Laraia MT,DeAndrade S.Continuum of care
for Alzheimer’s disease:a nurse education and counseling program.
Issues Ment Health Nurs 2001;22: 231–52.
48 Huang HL,Shyu YIL,Chen MC, Chen ST,Lin LC. A pilot study on a
home-based caregiver training program for improving caregiver
self-efficacy and decreasing the behavioralproblems of elders with
dementia in Taiwan.Int J Geriatr Psychiatry 2003;18: 337–45.
49 Magai C, Cohen CI,Gomberg D.Impact of training dementia caregivers
in sensitivity to nonverbalemotion signals.Int Psychogeriatr 2002;14:
25–38.
50 Finnema E,Droes RM,Ettema T,Ooms M, Ader H, Ribbe M, et al.
The effect of integrated emotion-oriented care versus usualcare on
elderly persons with dementia in the nursing home and on nursing
assistants:a randomized clinicaltrial. Int J Geriatr Psychiatry 2005;20:
330–43.
51 Camberg L,Woods P, Ooi WL, Hurley A,Volicer L,Ashley J, et al.
Evaluation of Simulated Presence:a personalized approach to enhance
well-being in persons with Alzheimer’s disease.J Am Geriatr Soc 1999;
47: 446–52.
52 Beck CK, VogelpohlTS, Rasin JH,Uriri JT, O’Sullivan P,Walls R,et al.
Effects of behavioralinterventions on disruptive behavior and affect in
demented nursing home residents.Nurs Res 2002;51: 219–28.
53 Long PW. When is a Difference Between Two Groups Significant?
Internet MentalHealth,2011 (http://www.mentalhealth.com/dis-rs/
rs-effect_size.html).
54 Algase DL,Beck C, KolanowskiA, WhallA, Berent S,Richards K,et al.
Need-driven dementia-compromised behavior:an alternative view of
disruptive behavior.Am J Alzheimers Dis Other Demen 1996;11: 10–9.
55 Kitwood T, Bredin K.Towards a theory of dementia care:personhood
and well-being.Ageing Soc 1992;12: 269–87.
56 Remington R.Calming music and hand massage with agitated elderly.
Nurs Res 2002;51: 317–23.
57 Weiner MF, Tractenberg RE,Sano M, Logsdon R,Teri L, Galasko D,et al.
No long-term effect of behavioraltreatment on psychotropic drug use for
agitation in Alzheimer’s disease patients.J Geriatr Psychiatry Neurol2002;
15: 95–8.
58 Eggermont LH,Blankevoort CG,Scherder EJ.Walking and night-time
restlessness in mild-to-moderate dementia:a randomized controlled trial.
Age Ageing 2010;39: 746–9.

The British Journalof Psychiatry (2014)
205, 436–442.doi: 10.1192/bjp.bp.113.141119
Data supplement
Table DS1 Interventions with evidence of usefulness
Study
Type of
intervention
Degree of
agitation of
participants
Quality
grade
Quality
score
Total
participants,
n
Therapeutic
regimen
Separate
controlgroup
Immediate
outcomea SES (95% CI)
Long-term
outcomea
SES
(95% CI)
Buettner
& Ferrario
(1997)23
Activities None 2b 7 66 30 weeks of neuro-
developmental
sequenced activities
(e.g. cooking group),
frequency unclear
Usualcare including
activities
Significantly improved
at 2 time-points but
not third
NC None NC
Cohen-Mansfield
et al(2006)27
Activities None 2b 6 105 5 sessions of activity
matched to self-identity
roles
Standard activities Significant
improvement
NC None NC
Kolanowskiet al
(2011)24
Individualised
activities
Some 1b 13.5 128 15 sessions activities
adjusted to functional
level(FL), personality
(PSI)or both (FL+PSI)
Standard activities NS FL 0.2 (70.3 to 0.7);
PSI 1.5 (0.9 to 2.0);
FL+PSI1.0 (0.4 to 1.5)
No difference
(1 week)
NC
Kovach et al
(2003)25
Activities Some 2b 7.5 78 Varied activities
matched to arousal
level(e.g. music,
exercise,storytelling)
Usualcare Significant
improvement in
visualanalogue
scale of agitation
NC None NC
Lin et al
(2009)22
Activities Some 2b 6 133 28 sessions of
Montessoriactivities
Presence (having
researcher present)
Significant
improvement
NC None NC
Cooke et al
(2010)30
Music therapy using
a specific protocol
Some 1b 11.5 47 Music therapy 3 times
a week for 8 weeks
Reading group Significant
improvement
70.9 (71.2 to 70.6) NC as
crossover
NC
Lin et al
(2011)28
Music therapy using
a specific protocol
None 2b 9.5 104 Music therapy twice
a week for 6 weeks
Usualcare Significant
improvement
70.6 (70.9 to 70.4) Significant
improvement
(1 month)b
70.6 (70.9
to 70.3)
Sung et al
(2012)29
Music therapy using
a specific protocol
Some 2b 10.5 55 Music therapy twice
a week for 6 weeks
Usualcare NS 70.5 (71.0 to 0.0) None NC
Hawranik
(2008)31
Sensory interventionsSome 2b 7 51 5 sessions of
therapeutic touch
on consecutive days
Placebo therapeutic
touch;usualcare
NS (totalagitation) NC NS NC
Lin et al
(2009)22
Sensory interventionsSome 2b 6 133 28 sessions of
acupressure over
4 weeks
Presence (researcher
present)
Significant
improvement
NC None NC
Remington
(2002)56
Sensory interventionsSome 2b 11 68 Hand massage,hand
massage and calming
music; given once
Usualcare Significant
improvement
Hand massage 70.6
(71.1 to 70.1);plus music
71.3 (71.9 to 70.8)
None NC
Van Weerte
(2005)34,c
Sensory interventionsNone 2b 10 125 Snoezelen over
18 months
Usualcare Significant
improvement in
aggression;
PN and VA NS
PN 70.1 (70.3 to 0.2);PA
71.4 (71.7 to 71.0);
VA 73.9 (74.4 to 73.4)
None NC
Woods et al
(2005)32
Sensory interventionsSome 2b 11 60 Therapeutic touch
twice a day for 3 days
Placebo massage:
usualcare
NS NC None NC
Woods et al
(2009)33
Sensory interventionsSome 2b 10 64 Therapeutic touch
twice a day for 3 days
Placebo therapeutic
touch or usualcare
NS NC NS 5 days later
(v. placebo or
usualcare)
NC
(continued)

2
Table DS1 Interventions with evidence of usefulness (continued)
Study
Type of
intervention
Degree of
agitation of
participants
Quality
grade
Quality
score
Total
participants,
n
Therapeutic
regimen
Separate
controlgroup
Immediate
outcomea SES (95% CI)
Long-term
outcomea
SES
(95% CI)
Chenoweth
et al(2009)35
Person-centred care
and communication
skills
Significant 1b 11.5 180 Training plus 2 site
visits and telephone-
based supervision
Usualcare Significant
improvement in
restraint use; PRN
and care recipient
quality of life NS
71.8 (71.9 to 71.6) Significant
improvement
(8 weeks)b
72.2 (72.4
to 72.0)
Deudon et al
(2009)36
Person-centred care
and communication
skills
Not specific 2b 7.5 306 Training including
issuing Staff Instruction
Cards on BPSD,
ongoing support
Usualcare Significant
improvement
70.32 (70.48 to 70.16) Significant
improvement (20
weeks)b
70.3 (70.5
to 70.2)
McCallion et al
(1999)38
Person-centred care
and communication
skills
Some 1b 10 66 Nursing assistants
delivered
7 communication-
focused sessions to
family caregiver,
with supervision
Usualcare Verbalagitation,
physical
non-aggression and
irritability improved;
aggression did not
PN 70.2 (70.6 to 0.1);
VA 70.1 (70.4 to 0.3);
PA 0.0 (70.3 to 0.4)
Only verbal
aggression and
irritability remained
significant
(3 months)b
PN 70.6
(71.0 to
70.3); VA
70.7 (71.0
to 70.3);
PA 0.1 (70.3
to 0.4)
McCallion et al
(1999)37
Person-centred care
and communication
skills
Some 2b 6 105 Communication skills
training with ongoing
support
Partialcrossover
– usualcare
Significant
improvement in
all agitation
70.4 (70.7 to 70.2) Significant
improvement;
physicalrestraints
improved, PRN
worsened (6 months)b
70.2 (70.5
to 0.1)
Sloane et al
(2004)39
Person-centred care
and communication
skills
None 2b 6 73 Trained in person-
centred bathing/towel
bath with support
implementing
Crossover – usual
care
Significant
improvement for
both showering
and towelbath
conditions
NC None NC
Chenoweth et al
(2009)35
Dementia care
mapping
Significant 1b 11.5 191 12 h dementia care
mapping plus support
implementing
Usualcare Significant
improvement.
PRN, quality of life
and restraint NS
71.4 (71.5 to 71.3) Significant improve-
ment. PRN and
restraint NS (4
months)b
71.5 (71.6
to 71.3)
BPSD,Behaviouraland PsychologicalSymptoms of Dementia; FL, (matched to) functionallevel; NC, not calculable; NS, not significant,PSI, (matched to) interest only;PA, physicalaggression; PN, physicalnon-aggression; PRN, prescription of drugs as required;
SES,standardised effect size; VA, verbalaggression/agitation.
a. In comparison with the controlcondition.
b. Originalpaper used a random effects model, a marginalmodelbased on generalised estimating equations or repeated-measures analysis of variance/covariance.
c. Majority but not allparticipants were randomised. We were unable to access originaldata.

Table DS2 Interventions without evidence of usefulness
Study
Type of
intervention
Degree of
agitation for
participation
in the study
Quality
grade
Quality
score
Total
patients,
n
Therapeutic
regimen
Separate
controlgroup
Immediate
outcome SES (95% CI)
Long-term
outcome
SES
(95% CI)
Ancoli-Israelet al
(2003)40
Light therapy Significant 2b 6 92 2 h daily light therapy
for 10 days (a.m.
or p.m.)
Placebo red light
during a.m.
Verbalagitation
worsened
72.0 (72.4 to 71.6) Nonea 70.3 (70.6
to 0.1)
Burns et al(2009)41 Light therapy Some 1b 12.5 48 2 h daily light therapy
for 2 weeks
Standard light NS 70.2 (70.6 to 0.2) NS (4 weeks)a 70.3 (70.7
to 0.2)
Dowling et al
(2007)42
Light therapy Some 2b 6 70 Activities in brightly lit
area (outside/lightbox)
1 h/day for 10 weeks
Similar activities
in a non-brightly
lit area
Significantly
worsened
p.m.light 4.0 (3.1 to 4.9);
a.m.light 7.0 (5.8 to 8.3)
(NPI)
None NC
Ballard et al
(2002)43
Aromatherapy Significant 2b 6 72 56 sessions of
Melissa oilmassage
Odourless sunflower
oil
Significant
improvement
NC None None
Burns et al
(2011)44
Aromatherapy Significant 1b 12.5 94 168 sessions aroma-
therapy massage (plus
placebo/donepezil)
Placebo aromatherapy
massage (plus
placebo/donepezil)
NS NC None None
Gormley et al
(2001)45
Training family
caregivers in BMT
Some 1b 11.5 65 4 sessions training
BMT
Given non-behavioural
advice and signposting
NR None NS agitation and
caregiver burden (2
weeks)a
70.6 (71.0
to 70.2)
(RAGE)
Teriet al(2000),
Weiner et al
(2002)46,57
(short- and
long-term effects)
Training family
caregivers in BMT
Significant 1b 11 77 11 sessions training
BMT
Placebo medication
(we did not consider
psychotropic
medication group)
NS (agitation,
caregiver burden)
NC NS (3,6 and 12
months)
NC
Huang et al
(2003)48
Training family
caregivers in CBT
Significant 2b 8 59 2 home and
13 telephone
consultations
Written educational
materials and social
telephone calls
Unclear as baseline/
change scores not
analysed;significantly
different at T2
70.3 (70.6 to 70.0) Unclear (3 months)70.2 (70.5
to 1.1)
Wright et al
(2001)47
Training family
caregivers in CBT
Significant 2b 7 93 3 home and
2 telephone
consultations training
CBT
Usualcare NS NC NS agitation,care-
giver wellbeing
(9 months)
NC
Eggermont et al
(2010)58
Exercise None 2b 6.5 112 30 sessions
of walking
Socialvisit,outside NS (restlessness) NC NS (7 weeks) NC
Finnema et al
(2005)50
Training programmes
for paid caregivers
without supervision
None 2b 8 146 Whole staff ethos
training,selected staff
intensive training, groups
and supervision on
emotion-oriented care
Usualcare NS NC None None
Magaiet al
(2002)49
Training programmes
for paid caregivers
without supervision
None 2b 6.5 91 Non-verbal
communication skills
training, no supervision
Educationaltraining
(placebo),usualcare
(control)
NS NC NS (9,12, 15 weeks)NC
(continued)

4
Table DS2 Interventions without evidence of usefulness (continued)
Study
Type of
intervention
Degree of
agitation for
participation
in the study
Quality
grade
Quality
score
Total
patients,
n
Therapeutic
regimen
Separate
controlgroup
Immediate
outcome SES (95% CI)
Long-term
outcome
SES
(95% CI)
Beck et al
(2002)52
Other:mix of ADL,
communication skills
and psychosocialactiv-
ities
Some 2b 7 96 60 sessions
promoting functional
independence,
psychosocial
Intervention or both
Socialcontact
(placebo),usual
care (control)
NS NC NS (1, 2 months) NC
Hong (2011)59 As above None 2b 7 55 Culturally familiar
environment from
youth with sensory
activities
Same familiar
environment but
no activities
NS 70.3 (70.7 to 0.1) None 2 NC
Camberg et al
(1999)51
Simulated presence Not specified 2b 8 54 Simulated presence
tape at least twice
a day while care
recipient agitated
Crossover – neutral
tape (placebo),
usualcare
NS (totalagitation) NC None NC
ADL, activities of daily living; BMT, behaviouralmanagement therapy; CBT, cognitive–behaviouraltherapy; CMAI, Cohen-Mansfield Agitation Inventory; NC, not calculable; NPI, Neuropsychiatric Inventory; NR, not reported; NS, not significant (P40.05); PRN,prescription
of drugs as required; RAGE, Rating Scale for Aggressive Behaviour in the Elderly; SES,standardised effect size.
a. Originalpaper used either a random effects model, a marginalmodelbased on generalised estimating equations or repeated-measures analysis of variance/covariance.

10.1192/bjp.bp.113.141119Access the most recent version at DOI:
2014, 205:436-442.BJP
Rumana Z. Omar, Cornelius Katona and Claudia Cooper
Gill Livingston, Lynsey Kelly, Elanor Lewis-Holmes, Gianluca Baio, Stephen Morris, Nishma Patel,
systematic review of randomised controlled trials
Non-pharmacological interventions for agitation in dementia:
Material
Supplementary http://bjp.rcpsych.org/content/suppl/2015/03/02/205.6.436.DC1.html
Supplementary material can be found at:
References http://bjp.rcpsych.org/content/205/6/436#BIBL
This article cites 56 articles, 10 of which you can access for free at:
permissions
Reprints/ [e-mail] to
To obtain reprints or permission to reproduce material from this paper, please
to this article at
You can respond /letters/submit/bjprcpsych;205/6/436
from
Downloaded The Royal College of PsychiatristsPublished by on July 7, 2015http://bjp.rcpsych.org/