SEARCH

SEARCH BY CITATION

Keywords:

  • music therapy;
  • systematic review;
  • narrative synthesis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

Objective

Recent reviews on music therapy for people with dementia have been limited to attempting to evaluate whether it is effective, but there is a need for a critical assessment of the literature to provide insight into the possible mechanisms of actions of music therapy. This systematic review uses a narrative synthesis format to determine evidence for effectiveness and provide insight into a model of action.

Method

The narrative synthesis framework consists of four elements: (i) theory development; (ii) preliminary synthesis of findings; (iii) exploration of relationships between studies; and (iv) assessment of the robustness of the synthesis.

Results

Electronic and hand searches identified 263 potentially relevant studies. Eighteen studies met the full inclusion criteria. Three distinctive strands of investigations emerged: eight studies explored behavioural and psychological aspects, five studies investigated hormonal and physiological changes, and five studies focused on social and relational aspects of music therapy. The musical interventions in the studies were diverse, but singing featured as an important medium for change.

Conclusions

Evidence for short-term improvement in mood and reduction in behavioural disturbance was consistent, but there were no high-quality longitudinal studies that demonstrated long-term benefits of music therapy. Future music therapy studies need to define a theoretical model, include better-focused outcome measures, and discuss how the findings may improve the well-being of people with dementia. Copyright © 2012 John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

Evidence for the benefits of music therapy in dementia remains inconclusive. Literature reviews on music therapy in dementia conducted to date (Brotons et al., 1997; Koger et al., 1999; Brotons, 2000; Vink et al., 2003, 2011; Ridder, 2005) have found short-term reductions in behavioural disturbance and improved mood, but evidence for long-term benefits is lacking (Livingston et al., 2005). Despite this, music therapy remains popular, and feedback from practitioners supports the idea that it has beneficial effects.

Meta-analyses of randomised controlled trials (RCTs) generally provide more reliable evidence in evaluating healthcare interventions (Evans, 2003). However, RCTs are not always the most suitable research design for psychosocial interventions because provisions of double blinding to treatment or placebo condition are not always practically possible or ethically suitable. Individual cases are explored in more detail in qualitative studies or in single-case studies, but these studies are automatically excluded from standardised quantitative meta-analysis. Although Vink et al. (2003, 2011) ‘accepted all behavioural and psychological tools reported by the authors of the identified primary studies’ for their Cochrane review on music therapy in dementia care, reported outcomes of RCTs may not always guarantee their validity if the outcome measures they used were not psychometrically validated. Traditional narrative reviews may offer more flexibility to accommodate various study designs; however, these reviews can be seen as less trustworthy if review methods such as inclusion and exclusion criteria or quality assessment of studies are not made explicit. There is a need to evaluate evidence from diverse music therapy studies in a more transparent, systematic manner.

Various forms of narrative synthesis (NS) are widely used in systematic literature reviews. However, NS has been criticised because of the lack of consensus on its constituent elements and conditions for establishing trustworthiness (Pope et al., 2007). Popay et al. (2006) devised a guide ‘to make the process of NS more systematic and to minimise bias’. NS is defined as ‘an approach to the systematic review and synthesis of findings from multiple studies that relies primarily on the use of words and text to summarise and explain the findings’. NS can still include the statistical analysis of the findings, but the key to this approach is not only to review what worked but also to investigate why and how an intervention might have worked. NS enables a systematic evaluation of both process-based and outcome-based studies. This approach is particularly relevant to a review on music therapy literature.

Objective

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

This study aims to conduct a NS systematic review of literature on music therapy in dementia using the four interactive NS elements: (i) theory development; (2) preliminary synthesis of findings; (iii) exploration of relationships between studies; and (iv) assessment of the robustness of the synthesis (Figure 1).

image

Figure 1. Narrative synthesis process (adapted from Popay et al., 2006).

Download figure to PowerPoint

Method (narrative synthesis Element 1: theory development)

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

Music therapy involves building a therapeutic relationship through listening and responding to a sound. Listening requires attention to an environment (Aldridge and Aldridge, 1992). Ridder (2004) argued that a client needs a moderate level of musical stimuli and that an ability to adjust one's response to external stimuli is linked with arousal regulation. Only after a balanced arousal level is achieved can the client form and sustain a musical dialogue with the therapist (Ridder, 2003, 2011; Ridder and Aldridge, 2005).

Music therapy methods and techniques used in clinical practice are diverse. Ansdell (1995) argued that an improvisation-based active music therapy model enables a therapist to attune to ‘the client's unique physical, mental and social condition at that time’, and in turn, this encourages the client to listen and be aware of self and others. The incorporation of familiar songs into improvisations is particularly common with clients with dementia. The use of a rigorously designed structured manual, including guidance in relation to improvisation, may improve the replicability of a music therapy intervention. This, however, should be used in the context of a therapeutic approach delivered by a trained music therapist.

Group music therapy may encourage social interaction between group members, thus reducing social isolation and assisting in communicating feelings and ideas (Aldridge, 1996). Fundamentally, music making is non-verbal, and this offers an alternative means for self-expression and communication when the conventional use of language becomes difficult. The effects of music on language skills have been explored in a number of studies (e.g. Prickett and Moore, 1991; Brotons and Koger, 2000; Suzuki et al., 2004), and clinicians often report clients' preserved memory of familiar songs. This suggests that music is an accessible medium for people with dementia. However, despite the wealth of music and music therapy literature in dementia care, there are no in-depth reviews exploring the mechanisms of music therapy interventions.

Search strategy

Electronic searches on MEDLINE, EMBASE, PsycINFO, CINAHL, Cochrane Library, Web of Science, Journal of Music Therapy, and Nordic Journal of Music Therapy were conducted. Search terms used alone or in combination were music therapy, music, dementia, Alzheimer's, limited to 1985–2011.

Criteria for considering studies for this review

Studies whose types of participants included people of any age and either gender and with dementia of any type living in both community and continuing-care settings were considered for this review.

Definition of music therapy for this review

  • Theory:
    • Use of music and its elements to promote clients' health and well-being
    • Use of a defined therapeutic model
  • Practice/operational definition:
    • Conducted by trained music therapists
    • Application of systematic and replicable therapy process (e.g. assessment, individual goal setting, treatment, evaluation)
  • Evidence:
    • Qualitative: reflexivity, clinical observation, and systematic evaluation of therapy process
    • Quantitative: pre-defined evaluation of musical/behavioural/relational components of treatment
  • Types of interventions:
    • Both active and receptive music therapies either in individual or group format were considered. Active music therapy applies to a model where both clients and therapists participate in music making. Receptive music therapy implies clients listening to recorded or live music that are selected to meet individual clinical needs.

Inclusion criteria

Quantitative, qualitative, and mixed-method studies that met the inclusion criteria were included for the review.

  • Standard inclusion criteria
    • Primary research relating to music therapy with people with dementia
    • Empirical studies in English language published in peer-reviewed journals between 1985 and 2011
    • Studies fulfilling the definition of music therapy
    • Studies clearly stating their aims, objectives, and methods
    • Studies clearly stating the types of music interventions and theoretical orientations used

In addition, the studies needed to fulfil the following inclusion criteria according to their study types:

  • Inclusion criteria for quantitative studies
    • RCTs, quasi-experimental studies including non-randomised controlled studies and before-and-after studies, and observational studies
    • Studies that use psychometrically robust outcome measures to collect primary outcome data
    • Studies that met at least 40% of Downs and Black's (1998) checklist criteria
  • Inclusion criteria for qualitative studies
    • Studies that clearly explain their choice of qualitative design
    • Studies whose analysis of qualitative data is clearly linked with study aims and objectives
    • Studies that met at least seven out of 10 Critical Appraisal Skills Programme (CASP) qualitative research appraisal criteria (Public Health Resource Unit, 2006)
  • Inclusion criteria for mixed-method studies
    • Studies whose choice of mixed methods is clearly explained
    • Studies whose psychometrically tested outcome measures are used to collect quantitative data
    • Studies whose synthesis of qualitative and quantitative data is clearly linked with study aims and objectives
    • Studies that met at least seven out of 10 CASP criteria.

Exclusion criteria

  • Studies with mixed populations (those with and without dementia) but do not differentiate the results between the two groups
  • Case report, conference paper, dissertation, personal opinion, and commentary
  • Non-empirical studies with the main focus on assessment of cognitive function, assessment of responses to pre-determined musical components, outcome measures development, philosophical discussion, or theory development
  • Studies combining music therapy with another intervention

Results (narrative synthesis Element 2: developing a preliminary synthesis)

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

The electronic searches identified 253 potentially relevant titles. Of these, 193 abstracts did not meet the inclusion criteria, leaving 60 abstracts for further investigation. Hand searches of music therapy journals (Journal of Music Therapy, British Journal of Music Therapy, and Australian Journal of Music Therapy) identified a further 10 studies. Full-text articles were obtained for all the 70 studies and were read through for further evaluation. Of these, 18 studies (15 quantitative studies and three qualitative/mixed-method (QMM) studies) met the full inclusion criteria and were included for this review (Figure 2).

image

Figure 2. Study selection process.

Download figure to PowerPoint

Fifty-two studies were excluded. Excluded studies comprised eight case reports, three discussion and theory development papers, four musical assessment on cognitive function, one musical assessment on language skills, eight assessment of response to pre-determined music components, two outcome measure development, three multi-modal interventions, two that included subjects without dementia, one conference paper, 14 music-based activity studies (e.g. comparison of music preference with other activities, music activities provided by activity coordinators or caregivers, music listening while bathing), and six that used non-psychometrically validated outcome measures to evaluate the main outcomes of their studies.

The 18 studies that met the full inclusion criteria were categorised according to the classification of study designs used in the ‘Systematic reviews: CRD's guidance for undertaking reviews in health care’ (Centre for Reviews and Dissemination (CRD), 2009). The following acronyms are used: RCTs (n = 6), non-randomised controlled studies (NRC) (n = 4), before-and-after studies (B&A) (n = 5), QMM (n = 3).

Quality assessment of the included studies

Two quality assessment tools were chosen on the basis of the favourable reviews in the ‘HTA Evaluating non-randomised interventions’ (Deeks et al., 2003) and the Centre for Reviews and Dissemination's guidance for systematic reviews (CRD, 2009). The Downs and Black (1998) checklist was used to assess the study quality of the 15 quantitative studies. Three QMM studies were evaluated using CASP (Public Health Resource Unit, 2006). Two reviewers, OM and NC, evaluated the studies independently to minimise bias; the results were validated by a third reviewer (MO). Any discrepancies in the scores were discussed and resolved in the review meetings.

Downs and Black (1998) checklist

The strength of this checklist includes its suitability to assess non-randomised studies as well as randomised studies. It consists of 27 items across five domains: reporting, external validity, internal validity (bias), internal validity (confounding), and power. Two items were omitted, as they were not applicable to psychosocial interventions. Two further items were only applicable to RCTs, and the item power was not used for this review. The maximum score varied depending on study designs: RCT = 25, NRC = 23, B&A with follow-up (F/U) = 19, and B&A without F/U = 17. Mean scores for each study category were as follows: RCT (n = 6) = 17.5 (70%), NRC (n = 4) = 13.8 (60%), B&A with F/U (n = 3) = 11.0 (58%), and B&A without F/U (n = 2) = 11.5 (68%). Overall mean was 65%.

  • Reporting: The aims of the studies, outcomes to be measured, interventions, and the main findings of the studies were reported adequately in over 85% of studies. None of the 15 studies reported potential adverse events that may be a consequence of music therapy intervention. Music therapy may be regarded as a non-pharmacological treatment, but adverse effects that could result from dealing with challenging psychological issues should perhaps be mentioned and procedures in such events be discussed.
  • External validity: Only the two studies by Raglio et al. (2010a, 2008) provided a full description of the selection procedure of study participants and clinical contexts (places, population, and facilities). Although 13 out of 15 studies described how participants from the study sites were selected, most studies did not provide sufficient information to judge if the participants were typically representative of this population. This meant that we were unable to determine whether the findings from these studies were generalisable.
  • Internal validity (bias): This was the strongest domain for all but one study (Groene, 1993). The compliance with the interventions was reliable (100%), and 14 studies (93%) provided sufficient evidence that appropriate statistical tests were used for analysis.
  • Internal validity (confounding): Only two studies (Guétin et al., 2009; Raglio et al., 2010a) provided a full explanation for adjustment of confounding. The conclusions of three studies were drawn from analysis of treatment rather than the results based on intention to treat. High dropout rates were found in five studies (range: 25–57%), some highlighting the challenge of collecting physiological data from this population such as saliva collection (Suzuki et al., 2004; Takahashi and Matsushita, 2006) and blood samples (Kumar et al., 1999).

Critical Appraisal Skills Programme (2006)

The three studies that met the CASP standard (Ridder and Aldridge, 2005; Ahonen-Eerikäinen et al., 2007; Ridder et al., 2009) gave clear explanations for their choice of QMM research design, conducted rigorous data analysis, and provided clear statements of the findings. Recruiting procedure and inclusion/exclusion criteria of the study subjects were not fully described in two studies (Ridder and Aldridge, 2005; Ridder et al., 2009), although the full information on these studies was provided elsewhere (Ridder, 2003). One study (Ahonen-Eerikäinen et al., 2007) gave a full description of the procedure for data collection and analysis but offered a limited explanation of the music therapy intervention itself.

Characteristics of included studies

All study subjects, except participating family carers (n = 29), had a diagnosis of moderate to severe dementia (n = 589). Quantitative and mixed-method studies aimed to assess effects of music therapy on the following: (i) behavioural and psychological symptoms; (ii) physiological changes; or (ii) carer/care receiver relationships. Qualitative elements of the studies focused on exploring the effects of music therapy on the quality of life of people with dementia. Seventeen studies used an active music therapy model. An improvisation-based model was favoured by Raglio et al. (2010a, 2010b, 2008); but group music therapy programmes based on pre-planned music activities appeared more common (n = 11). The three QMM studies explored the clinical significance of therapist/client relationship through interactive music making. The summary of each study is provided in Tables 1 and 2. For a comparison of effect sizes of the studies, the results were converted to Cohen's d where possible.

Table 1. Summary of quantitative studies
Author (year)CountryStudy designNumber of participants (women)Age, mean/range (years)Focus of studyMain outcome measuresSummary of interventionData collectionNumber of dropoutsOutcome of treatmentEffects maintained at F/U? (yes/no)Quality (D&B 1998)
ExperimentalControl
  1. RCT, randomised controlled trials; NRC, non-randomised controlled studies; B&A, before-and-after studies; E, experimental group; C, control group; B/L, baseline; F/U, follow-up; SC, standard care; D&B, Downs and Black checklist score; high, high-quality study (above 81%); medium, medium-quality study (66–80%); fair, fair-quality study (51–65%); low, low-quality study (below 50%); BEHAVE-AD, Behavior Pathology in Alzheimer's Disease Rating Scale; BP, blood pressure; BPSD, behavioural and psychological symptoms of dementia; CBQ, Caregiver Burden Questionnaire; CgA, salivary chromogranin A; CHF, congestive heart failure; CMAI, Cohen–Mansfield Agitation Inventory; DBRS, Disruptive Behavior Rating Scales; GDS, Geriatric Depression Scale; Hamilton, Hamilton scale; HR, heart rate; HRV, heart rate variability; MBBS, Montgomery and Borgatta Burden Scale; MMSE, Mini mental State Examination; MOSES, Multidimensional Observation Scale for Elderly Subjects; MT, music therapy; MTCS, Music Therapy Coding Scheme; NPI, Neuropsychiatric Inventory; PANAS, Positive and Negative Affect Scale; STAI, State–Trait Anxiety Inventory.

  2. a

    Addendum to the original statistical data published in 2012.

Raglio et al. (2010a)aItalyRCT60 (55)85 (E), 85 (C)Reduction of BPSDNPIImprovisation-based group MT: three 30-min session/week for 4 weeks; three cycles with a 1-month break between cyclesEducational and entertainment activities (SC)B/L and post-MT; F/U 1 month later7Improved (F = 4.09, p = 0.049)NoHigh (88%), 22/25
Raglio et al. (2010b)ItalyRCT20 (15)84 (E), 87 (C)Changes in physiological parametersNPI, HR, HRVImprovisation-based MT: two 30-min session/week for 15 weeksEducational and occupational activities (SC)B/L and post-MT; no F/U0HR and NPI: no change; NPI sub-score ‘depression’: improved (d = 0.43, p = 0.021); HRV: improved in 50% of MT group (χ2 = 6.1, p = 0.013)N/AFair (60%), 15/25
Guétin et al. (2009)FranceRCT30 (22)85 (E), 87 (C)Anxiety and depressionHamilton, GDS16 weekly individual receptive MT: listening to computer-programmed 20-min sequence of musicRest and readingB/L, Weeks 4 and 8, post-MT; F/U 8 weeks later2 (E), 4 (C)Hamilton scale: improved (E: d = 2.98; C: d = 0.05, p = 0.001); GDS: improved (E: d = 1.57; C: d = 0.09, p = 0.002)Hamilton: yes (E: d = 1.96; C: d = 0.11, p ≤ 0.001); GDS: yes (E: d = 0.67; C: d = −0.04; p = 0.003)High (88%), 22/25
Raglio et al. (2008)ItalyRCT59 (50)84(E), 86(C)Reduction of BPSDNPI, MTCSImprovisation-based group MT: 30 min × 10 sessions × three cycles over 16 weeksEducational and entertainment activities (SC)B/L, Week 8, post-MT; F/U 4 weeks later2NPI: improved (F = 5.06; p < 0.002); MTCS: empathetic behaviour increased (d = 0.61) and non-empathetic behaviour decreased (d = 1.8) (E)NPI: yes (F = 12.65; p = 0.0007)High (92%), 23/25
Svansdottir and Snaedal (2006)IcelandRCT4671–87Reduction of BPSDBEHAVE-ADThree 30-min group MT/week for 6 weeks; singing songs with guitar accompanimentStandard careB/L, post-MT; F/U 4 weeks later8Total score: no change; subscale ‘activity disturbances’ (Ad): decreased (p = 0.02); total subscales of ‘aggressiveness’, ‘anxiety’, and ‘Ad’: decreased (p < 0.01)NoLow (48%), 12/25
Groene (1993)USARCT30 (16)78Cognition, wanderingMMSESeven daily 1:1 sessions: five MT and two reading or two MT and five reading, ‘play-along’ with recorded musicDaily, no F/U0MMSE: no change; wandering behaviour: no changeN/ALow (44%), 11/25
Okada et al. (2009)JapanNRC8784 (E), 81 (C)Autonomic nervous system and CHF eventsHRV, CHF eventsWeekly 45-min group MT, use of well-known songsHRV pre-MT, during, and post-MT on the same day, blood sample pre-MT and post-MTNot discussedHRV (rMSSD): improved during MT (d = −0.55) and improved further post-MT (d = −0.20); CHF events: less frequent for MT groupN/AFair (57%), 13/23
Ledger and Baker (2007)AustraliaNRC45 (40)85(E), 86(C)Long-term effects on agitationCMAIWeekly group MT for 42+ weeks/year, listening to live music; song requests and discussionsFive time points over 1 year15 (60 initially recruited)Short-term reduction of agitation (F = 2.61, p < 0.05)N/AMedium (70%), 13/23
Takahashi and Matsushita (2006)JapanNRC43 (33)83Long-term effectsCortisol level, BPWeekly active reminiscence group MT over 2 yearsB/L, 6 months, 1 and 2 years6 (E), 5 (C)Systolic blood pressure: increased more in control group (d = −1.08); cortisol level: no differenceN/AFair (61%), 14/23
Suzuki et al. (2004)JapanNRC23 (15)83 (E), 85 (C)Behaviour, endocrine functionsMMSE, MOSES, CgATwo sessions/week for 8 weeks, used Clair and Bernstein's MT protocolB/L, Weeks 4 and 8; no F/U4 (E), 6 (C)MMSE language subscale: improved (d = −0.60, p = 0.012); MOSES: ‘irritability’ sub-score: improved (d = 0.68, p = 0.0001); CgA: improved (d = 0.72, p = 0.048)N/AFair (52%), 12/23
Brotons and Marti (2003)USAB&A28 (14)76 patients, 73 carersPatients with Alzheimer's and their familiesNPI, CMAI, STAI-SResidential group MT project for 14 carers and 14 patients over 12 days; 10 sessions for patients, seven joint sessions, and four sessions for carers onlyPre-test, Day 10; F/U 2 months laterUnclearNPI: improved (χ2 = 17.72, p = 0.001); CMAI: improved (χ2 = 11.45, p = 0.003); STAI-S: improved (χ2 = 14.72, p = 0.001)NoFair (58%), 11/19
Ashida (2000)USAB&A20 (17)86Depressive symptomsCornell Scale for DepressionWeeks 1 (B/L) and 2: no MT, Week 3: daily group reminiscence MT, Week 4: no MTB/L, end of Weeks 2, 3, and 40Improved (p < 0.05)YesLow (47%), 9/19
Kumar et al. (1999)USAB&A2078Hormonal changes and psychological functionsMelatonin, norepinephrine, epinephrineGroup MT five times/week for 4 weeks (total = 20), used Clair and Bernstein's protocolBlood sample: B/L, post-MT; F/U 6 weeks later7Melatonin level: improved (d = −0.73, p = 0.03); norepinephrine level: improved (d = −0.24, p = <0.0001); epinephrine level: improved (d = −0.27, p ≤ 0.0001)Melatonin level: improved further (d = −2.07 from B/L, p = 0.0001); norepinephrine and epinephrine levels: noMedium (68%), 13/19
Clair and Ebberts (1997)USAB&A30 (15)Not reportedEngagement between family membersMBBS, PANASGroup MT for people with dementia and caregivers, twice/week for 4 weeks, sessions based on Clair and Bernstein's protocolPre-MT and post-MT scores by carersNot reportedCaregivers' depression, burden, and self-reported health: no changesN/AFair (65%), 11/17
Brotons and Pickett-Cooper (1996)USAB&A47 (17)83AgitationDBRS30-min twice/week group MT consisting of various music activities,. total of five sessionsPre-MT and post-MT scores by carers27 (excluded from analysis)Agitation improved (d = 1.41, F = 16.33, p = 0.001); significant reduction in pacingN/AMedium (71%), 12/17
Table 2. Summary of mixed-method and qualitative studies
Author (year)CountryStudy designParticipants (women)Age, mean/range (years)Focus of studyMain outcome measuresSummary of interventionData collectionOutcomes of treatmentQuality (CASP 2006)
  1. MM, mixed-method study; BPSD, behavioural and psychological symptoms of dementia; CMAI, Cohen–Mansfield Agitation Inventory; NPI, Neuropsychiatric Inventory; MT, music therapy; QUAL, qualitative study; CASP, Critical Appraisal Skills Programme; ADRQL, Alzheimer's Disease-Related Quality of Life.

Ridder et al. (2009)DenmarkMM2 (1)Not reportedQuality of life, reduction of BPSDADRQL, CMAI, NPI, video analysisIndividual MT, four session/week for 4 weeks (total = 16), emphasis on an individualised use of songsADRQL, CMAI, NPI collected the week before and after MT, life story, video recordings of MTMr A: musical engagement and enjoyment, awareness of therapist, increased confidence, the effects of songs on verbal stimulation evident from video analysis, ADRQL and CMAI also improved; Mrs B: limited musical responses shown but was calmer post-MT and showed emotional response when personally meaningful songs were played, reference made to ‘singing lullabies’, ADRQL worsened possibly because of increased ‘isolation and remoteness’, CMAI and NPI distress scores improvedSampling: reasons for the choice of two subjects not explicit but explained in the full research reports available online (Ridder, 2003) (9/10)
Ahonen-Eerikäinen et al. (2007)CanadaQUAL6 (6)77–90Quality of life, peer interactionTherapists' notes, feedback from participants12 group MT over 4 months, no further details providedTherapists' field notes, session comments by two observers, video recordingHierarchy of descriptive categories: (i) MT allowed them to work through negative feelings; (ii) MT provided empowering experiences; and (iii) MT gave them ‘the joyful moments’ such as feelings of fun, energy, excitement, trust, and relaxation; the importance of peer interaction and peer supportData collection: contents of music therapy sessions not fully explained (9/10)
Ridder and Aldridge (2005)DenmarkMM1 (1)Not reportedRole of therapeutic singingVideo analysis, CMAI, heart rate, therapists' notes20 daily individual MT focus on building a therapist/client relationship through singingHeart rate measured pre-MT, during, and post-MT, video recording of each sessionMrs F: initially extremely restless and repeated the word ‘Bingo’ over 300–500 times during a session, therapy focused on the use of well-known songs, validation of her mood, and arousal regulation; as sessions progressed, Mrs F began to pause, make eye contact, and reach out for the therapist, reduction in repetitive ‘Bingo’, reduction in heart rate and in the use of antipsychotic medicationSampling: inclusion/exclusion criteria not fully explained (9/10)

Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

Characteristics of the studies according to study design

Randomised controlled trials (n = 6)

Five studies aimed to investigate the effects of music therapy on behavioural and psychological symptoms. The three Downs and Black ‘high-quality studies’ were all RCTs (Raglio et al., 2008; Guétin et al., 2009; Raglio et al., 2010a), but the two out of three ‘low-quality studies’ were also RCTs (Groene, 1993; Svansdottir and Snaedal, 2006).

Non-randomised controlled studies (n = 4)

The three studies that were conducted in Japan measured the physiological changes following group music therapy. The difficulties of collecting physiological samples (e.g. saliva and blood) resulted in high dropout rates. The weakest aspects of these studies were the lack of information on those who dropped out and the lack of clarity on any subsequent adjustments for the analysis of the study results. The authors also did not discuss further the importance of their results in the context of clinical practice.

Before-and-after studies (n = 5)

This study design was more frequent in the earlier studies (1996–2003) conducted in the USA. These studies had a trend towards using a highly structured pre-planned music therapy programme. Two studies focused on care receiver/caregiver relationships, two studies investigated changes in psychological symptoms, and one study investigated physiological changes following music therapy. The study quality varied greatly (Downs and Black score: range 47–71%).

Qualitative and mixed-method studies (n = 3)

Two mixed-method studies (Ridder and Aldridge, 2005; Ridder et al., 2009) used qualitative data to investigate the therapy process and the development of therapist/client relationships with the additional physiological and psychological quantitative data. The qualitative study by Ahonen-Eerikäinen et al. (2007) coded and categorised the verbal feedback from the group participants to obtain their perspectives: the flexible design offered an insight into the study participants' personal experiences of music therapy.

Focus of the investigations and study effects

This section attempts to categorise the studies according to their aims and objectives and then ‘identify any factors that might explain differences in direction and size of effect across the studies’ (Popay et al., 2006) and ‘how and why’ the interventions might have or have not had an effect. Within the 18 studies, three distinctive strands emerged.

  1. Behavioural and psychological aspects: effects on behavioural and psychological symptoms such as general behavioural and psychological symptoms of dementia (n = 3), agitation (n = 2), anxiety/depression (n = 2), and wandering (n = 1; total n = 8)
  2. Physiological aspects: effects on physiological changes such as heart rate, cortisol levels, and serum levels of melatonin (n = 5)
  3. Social and relational aspects: effects on social relationships (n = 5)
Effects of music therapy on behavioural and psychological symptoms

Over 50% of the quantitative studies (n = 8) used reduction of behavioural and psychological symptoms as the main outcome measure to assess the effects of music therapy interventions. The three studies by Raglio et al. provided modest but consistent evidence that improvisation-based small-group music therapy was effective in reducing neuropsychiatric symptoms. The effects were also maintained at follow-ups in one study (Raglio et al., 2008). The authors argued that non-verbal therapist/patient attunement and musical self-expression help organise and regulate the patients' behaviours and emotions. However, there was no attempt to investigate the specific aspects of their interventions that enabled the reduction of these symptoms. The control groups were offered educational, occupational, or entertainment activities that were part of standard care. Therefore, it is difficult to determine whether other forms of therapeutic interventions may also have a similar effect on the symptoms or these changes are specific to music therapy. Evidence on increased positive behaviour was less consistent. However, this may be partially due to the choice of study design (RCT) and the lack of suitable outcome measures that are sensitive to changes in positive behaviours.

Reduction of short-term agitation was observed in both quantitative and mixed-method studies (Brotons and Pickett-Cooper, 1996; Brotons and Marti, 2003; Ridder and Aldridge, 2005; Svansdottir and Snaedal, 2006; Ledger and Baker, 2007; Ridder et al., 2009). Brotons and Pickett-Cooper (1996) offered short-term intervention (five sessions) and found that it was still effective in reducing agitated behaviour and pacing. However, this study had a particularly high number of participants excluded from analysis (20 included for analysis out of 47 participants); therefore, the degree of evidence is questionable. Reduction in wandering behaviour was also reported by Groene (1993); however, the study quality was too low to draw any conclusion. Ledger and Baker (2007) attempted to investigate a long-term effect of music therapy on agitation; but this could not be established because of the large variation between the treatment and control groups and fluctuating Cohen–Mansfield Agitation Inventory scores.

Significant reductions in levels of anxiety and depression following ‘receptive relaxation music therapy’ were reported by Guétin et al. (2009). Four types of computer-programmed ‘musical sequence’ were designed for the study. The study did not discuss the role of the music therapist in detail, other than providing a brief description of ‘psychotherapist type of therapeutic relationships’ and ‘listening to the patient’ for an extended period. Ashida (2000) also found that the small-group intervention focusing on reminiscence using familiar songs was effective in reducing depressive symptoms. There was a significant improvement in the Cornell Depression Scale scores; however, provision of daily music therapy is not a common practice in most clinical settings. Hence, the positive results might not have been typical.

Effects of music therapy on physiological changes

Six studies (RCT = 1, NRC = 3, B&A = 1, and QMM = 1) examined hormonal and physiological changes observed following music therapy. Kumar et al. (1999) hypothesised that ‘musically medicated stimulation of several neurohormonal and neurotransmitter systems can be accompanied by behavioral changes’. Four studies used highly structured music programmes with an emphasis on group singing. Apart from mentioning the use of the Clair and Bernstein protocol (Clair and Bernstein, 1990) (Suzuki et al. and Kumar et al.), explanations of the music therapy intervention were minimal. The discussion of the clinical implications of their results and future research recommendations was also limited in contrast to the detailed report of hormonal and physiological changes. High dropout rates during sample collections also affected the quality of the studies' statistical analysis.

Despite these weaknesses, physiological changes related to music therapy were evident. Improvement in heart rate variability was reported by Raglio et al. (2010b) and Okada et al. (2009), and decreased heart rate was also documented (Ridder and Aldridge, 2005; Okada et al., 2009). Increased melatonin concentration in serum was associated with a ‘calmer mood’ amongst the patients (Kumar et al., 1999), and reduction of stress hormone (Salivary chromogranin A) was observed in Suzuki's (2003) study.

Effects of music therapy on social relationships

The RCT by Raglio et al. (2008) measured the therapist/client relationship using some items of the Music Therapy Coding Scheme, which was based on a psychodynamic framework (Raglio et al., 2006). Empathetic behaviour increased in the experimental group (Cohen's d = 0.61), and non-empathetic behaviour also decreased (Cohen's d = 1.8).

Two before-and-after studies (Clair and Ebberts, 1997; Brotons and Marti, 2003) explored effects of music therapy on the relationships between family caregivers and their family members with dementia. The interventions consisted of a series of structured music activities, but theoretical explanations for the choice of these particular activities were not provided for either studies. Brotons and Marti found a decrease in patients' neuropsychiatric symptoms, and carers reported improvement in social and emotional aspects of the patients. Carers' satisfaction with visits and increased carer engagement behaviour were observed in Clair and Ebberts' study. No significant statistical changes in carers' scores may reflect weak study designs, but these studies also highlighted the challenge of finding appropriate outcome measures for caregiver well-being following music therapy.

The three QMM studies investigated the social aspects of music therapy and the relationships between the clients and the therapists. The strengths of these studies included multi-dimensional analysis of the individuals' therapy process and therapy mechanisms as well as therapy outcomes. Ahonen-Eerikäinen et al. (2007) used ‘the qualitative paradigm, adapted grounded theory and narrative inquiry’ specifically to explore the effect of group music therapy on quality of life for clients with dementia. However, the content of music intervention was not discussed, even though detailed narrative summaries of clients' music therapy experiences have been provided. Discussion on the three-stage hierarchical music therapy process was provided: firstly, working through negative feelings; secondly, experience of empowerment; and thirdly, shared experiences of ‘the joyful moments’ such as energy, enjoyment, trust, and relaxation. The mixed-method studies by Ridder and Aldridge (2005) and Ridder et al. (2009) provided a detailed discussion on the role of therapeutic singing with three clients, which is based on the concepts of intrinsic communicative musicality (Trevarthen, 1999; Trevarthen and Malloch, 2008). The development of therapeutic relationships was based on musical engagement, validation of mood and emotional responses, and the clients' increased awareness of the therapist. A decrease in behavioural disturbance was observed. The findings from the QMM studies suggest that music therapy allows a therapist to meet the needs of individual clients regardless of the severity of dementia and to offer an individually tailored intervention.

Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

This is the first NS systematic review on music therapy literature. The use of defined inclusion/exclusion criteria, the application of a rigorous search strategy, and the quality assessment of the studies and systematic analysis of the findings made this review transparent. However, there are many areas that need to be improved for a future NS review.

Differentiating between music therapy and music activities studies is not straightforward. There is debate about whether music therapy depends on both theoretical and applied understanding or whether in addition it also requires a therapeutic relationship between therapist and client (Gold, 2009; Gold et al., 2011). For example, Gold et al. (2011) argued that a recent major trial of ‘receptive music therapy’ in depression (Brandes et al., 2010) was not music therapy but simply music listening because music therapy involved a therapeutic relationship between the therapist and the client (Gold, 2009). Raglio and Gianelli (2009) compared ‘music activities’ and ‘music therapy’ interventions in dementia and concluded that ‘music therapy requires clear theoretical and applicative bases along with methodological and scientific rigor in research design’. In addition, some empirical studies conducted by music therapists were music activity studies, and the concept of music therapy also differed according to cultural contexts. These factors made defining the inclusion/exclusion criteria for this review very challenging. There may be other empirical studies that meet the current inclusion criteria but were not included. The inclusion/exclusion criteria may also need to be revised for a future review.

The use of the Downs and Black checklist was valuable for the standardised assessment of randomised and non-randomised studies. However, the items are not weighted, and answers are given as either yes or no. This meant we could not differentiate the items that had a greater impact on the quality of the studies nor indicate the degree of the presence or of the lack of information provided. The RCT by Guétin et al. scored high in the Downs and Black assessment because of the clarity of the study design and procedure. However, the information on the role of music therapists was extremely limited; hence, the validity of the recorded music sequences as a ‘music therapy’ intervention rather than as a ‘music activity’ was questionable.

Singing featured as a medium for change in seven quantitative and qualitative studies. However, it was not possible to consolidate enough evidence to develop a new theory because of the heterogeneity of the musical interventions of different qualities, the diversity of research designs, and the limited explanations of the study findings and clinical implications.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References

This review is in agreement with previous systematic reviews (e.g. Livingston et al., 2005; Vink et al., 2003, 2011) that found evidence for short-term improvements in behavioural and psychological disturbance. There were no high-quality longitudinal studies that demonstrated how and why music therapy might have worked.

The limited availability of high-quality studies and the lack of evidence for long-term benefits of music therapy also highlight the difficulty of finding appropriate outcome measures to evaluate a complex intervention for people with dementia and the need for more research in this area. The choice of measurement instruments is crucial in assessing the long-term effects of an intervention. However, there may be a danger of choosing well-established measures in an attempt to provide trustworthy evidence, yet that may not be most relevant to the experiences of people with dementia. Mozley et al. (1999) found that a high proportion of older people with significant cognitive impairments could provide their own views on quality of life. Woods (2001) also argued that we need to ‘ask the person him/herself, before moving to proxy and observational measures’. Rigorous evaluations of qualitative data from music therapy sessions that show responses from individuals may particularly be important to investigate what aspects of music therapy people with dementia themselves may find meaningful.

Another issue of evaluating the long-term benefits of music therapy relates to the nature of dementia. The majority of the studies in this review were with people with moderate to severe dementia in residential settings. It may be important for a future music therapy study to re-define what will be the realistic and clinically relevant goal for this population on the basis of a clear theoretical framework. Mittelman (2008) argued, ‘perhaps changes cannot be sustained as the dementia becomes more severe. The intervention might still be deemed worthwhile if it improved the person's quality of life, even temporarily’. If the changes are so small or not sustainable because of the nature of dementia, it is even more vital to use a study design that is sensitive to change. It is only when the study findings are implemented in practice that the potential long-term effects of an intervention become beneficial for people with dementia.

Key points

  • Narrative synthesis systematic reviews enable a systematic evaluation of both process-based and outcome-based studies. This approach is particularly relevant when reviewing a complex intervention.
  • There was consistent evidence for short-term improvement in behaviour and mood following music therapy. Singing appeared to be associated with benefits in both quantitative and qualitative studies.
  • There was little evidence for longer-term benefits and no longitudinal studies investigating how and why music therapy might work.
  • Future studies on music therapy in dementia need to look beyond the potential short-term effects, employ a defined theoretical model, and use study designs with more appropriate outcome measures.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Objective
  5. Method (narrative synthesis Element 1: theory development)
  6. Results (narrative synthesis Element 2: developing a preliminary synthesis)
  7. Discussion (narrative synthesis Element 3: exploring the relationships within and between the studies)
  8. Limitations of this review (narrative synthesis Element 4: assessment of the robustness of the synthesis)
  9. Conclusion
  10. Conflicts of interest
  11. References
  • Ahonen-Eerikäinen H, Rippin K, Sibille N, Koch R, Dalby DM. 2007. “Not bad for an 85-year-old!”—the qualitative analysis of the role of music, therapeutic benefits and group therapeutic factors of the St. Joseph's Alzheimer's Adult Day Program Music Therapy Group. Canadian Journal of Music Therapy 13(2): 3762.
  • Aldridge D. 1996. Music therapy with the elderly. In Music Therapy Research and Practice in Medicine. From Out of the Silence. Jessica Kingsley Publishers: London; 186209.
  • Aldridge D, Aldridge G. 1992. Two epistemologies: music therapy and medicine in the treatment of dementia. The Arts in Psychotherapy 19: 243255.
  • Ansdell A. 1995. Improvising. In Music for Life. Jessica Kingsley Publishers: London; 2127.
  • Ashida S. 2000. The effect of reminiscence music therapy sessions on changes in depressive symptoms in elderly persons with dementia. J Music Ther 37: 170182.
  • Brandes V, Terris D, Fischer C, et al. 2010. Receptive music therapy for treatment of depression: a proof-of-concept study and prospective controlled clinical trial of efficacy. Psychother Psychosom 79: 321322.
  • Brotons M. 2000. An overview of the music therapy literature relating to elderly people. In Music Therapy in Dementia Care, Aldridge A (ed.). Jessica Kingsley Publishers: London; 3362.
  • Brotons M, Koger SM. 2000. The impact of music therapy on language functioning in dementia. J Music Ther 37: 183195.
  • Brotons M, Marti P. 2003. Music therapy with Alzheimer's patients and their family caregivers: a pilot project. J Music Ther 40: 138150.
  • Brotons M, Pickett-Cooper P. 1996. The effects of music therapy intervention on agitation behaviors of Alzheimer's disease patients. J Music Ther 33: 218.
  • Brotons M, Koger SM, Pickett-Cooper P. 1997. Music and dementias: a review of literature. J Music Ther 34: 204245.
  • Centre for Reviews and Dissemination. 2009. Systematic reviews: CRD's guidance for undertaking reviews in health care. Research Report, CRD, York.
  • Clair A, Bernstein B. 1990. A preliminary study of music therapy programming for severely regressed persons with Alzheimer's-type dementia. J Appl Gerontol 9(3): 299311.
  • Clair A, Ebberts A. 1997. The effects of music therapy on interactions between family caregivers and their care receivers with late stage dementia. J Music Ther 34: 148164.
  • Deeks JJ, Dinnes J, D'Amico R, et al. 2003. Evaluating non-randomised intervention studies. Health Technol Assess 7(27): 1179.
  • Downs SH, Black N. 1998. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health 52(6): 377384.
  • Evans D. 2003. Hierarchy of evidence: a framework for ranking evidence evaluating healthcare interventions. J Clin Nurs 12(1): 7784.
  • Gold C. 2009. All those things with music. Nordic Journal of Music Therapy 18(1): 12.
  • Gold C, Erkkilä J, Bonde LO, et al. 2011. Music therapy or music medicine? Psychother Psychosom 80: 304.
  • Groene R. 1993. Effectiveness of music therapy 1: 1 intervention with individuals having senile dementia of the Alzheimer's type. J Music Ther 30: 138157.
  • Guétin S, Portet F, Picot M, et al. 2009. Effect of music therapy on anxiety and depression in patients with Alzheimer's type dementia: randomised, controlled study. Dement Geriatr Cogn Disord 28(1): 3646.
  • Koger SM, Chapin K, Brotons M. 1999. Is music therapy an effective intervention for dementia? A meta-analytic review of literature. J Music Ther 36: 215.
  • Kumar A, Tims F, Cruess D, et al. 1999. Music therapy increases serum melatonin levels in patients with Alzheimer's disease. Altern Ther Health Med 5(6): 4957.
  • Ledger A, Baker F. 2007. An investigation of long-term effects of group music therapy on agitation levels of people with Alzheimer's disease. Aging Ment Health 11(3): 330338.
  • Livingston G, Johnston K, Katona C, Paton J, Lyketsos CG. 2005. Systematic review of psychological approaches to the management of neuropsychiatric symptoms of dementia. Am J Psychiat 162(11): 19962021.
  • Mittelman M. 2008. Psychosocial intervention research: challenges, strategies and measurement issues. Aging Ment Health 12(1): 14.
  • Mozley CG, Huxley P, Sutcliffe C, et al. 1999. ‘Not knowing where I am doesn't mean I don't know what I like’: cognitive impairment and quality of life responses in elderly people. Int J Geriatr Psychiatry 14: 776783.
  • Okada K, Kurita A, Takase B, et al. 2009. Effects of music therapy on autonomic nervous system activity, incidence of heart failure events, and plasma cytokine and catecholamine levels in elderly patients with cerebrovascular disease and dementia. Int Heart J 50(1): 95110.
  • Popay J, Roberts H, Sowden A, et al. 2006. Guidance on the conduct of narrative synthesis in systematic reviews. A product from the ESRC Methods Programme. Final Report, Lancaster University.
  • Pope C, Mays N, Popay J. 2007. Synthesizing Qualitative and Quantitative Health Evidence: A Guide to Methods. Open University Press: Maidenhead.
  • Prickett CA, Moore RS. 1991. The use of music to aid memory of Alzheimer's patients. J Music Ther 28: 101110.
  • Public Health Resource Unit. 2006. Critical Appraisal Skills Programme (CASP) 10 Questions to Help You Make Sense of Qualitative Research. Public Health Resource Unit: England.
  • Raglio A, Gianelli, M. 2009. Music therapy for individuals with dementia: areas of interventions and research perspectives. Curr Alzheimer Res 6: 293301.
  • Raglio A, Traficante D, Oasi O. 2006. A coding scheme for the evaluation of the relationship in music therapy sessions. Psychol Rep (99): 8590.
  • Raglio A, Bellelli G, Traficante D, et al. 2008. Efficacy of music therapy in the treatment of behavioral and psychiatric symptoms of dementia. Alzheimer Dis Assoc Disord 22(2): 158162.
  • Raglio A, Bellelli G, Traficante D, et al. 2010a. Efficacy of music therapy treatment based on cycles of sessions: a randomised controlled trial. Aging Ment Health 14(8): 900904.
  • Raglio A, Oasi O, Gianotti M, et al. 2010b. Effects of music therapy on psychological symptoms and heart rate variability in patients with dementia. A pilot study. Curr Aging Sci 3: 242246.
  • Raglio A, Bellelli G, Traficante D, et al. 2012. Addendum to ‘Efficacy of music therapy treatment based on cycles of sessions: a randomised controlled trial’ (Raglio et al., 2010). Aging Ment Health 16(2): 265267.
  • Ridder HM. 2003. Singing dialogue. Music therapy with persons in advanced stages of dementia. A case study research design. Doctorate thesis, Aalborg University, Denmark.
  • Ridder HM. 2004. When dialogue fails. Music therapy with elderly with neurological degenerative diseases. Music Therapy Today (online) 5(4). available at http://www.wfmt.info/Musictherapyworld/modules/mmmagazine/issues/20040727092613/20040727093215/MTT5_4_Ridder.pdf
  • Ridder HM. 2005. An overview of therapeutic initiatives when working with persons suffering from dementia. In Music Therapy and Neurological Rehabilitation. Performing Health, Aldridge D (ed.). Jessica Kingsley Publishers: London; 6182.
  • Ridder HM. 2011. How can singing in music therapy influence social engagement for people with dementia? Insights from the polyvagal theory. In Voicework in Music Therapy.: Research and Practice, Baker F, Uhlig S (eds). Jessica Kingsley Publishers: London; 130146.
  • Ridder HM, Aldridge D. 2005. Individual music therapy with persons with frontotemporal dementia: singing dialogue. Nordic Journal of Music Therapy 14(2): 91106.
  • Ridder HM, Wigram T, Ottesen AM. 2009. A pilot study on the effects of music therapy on frontotemporal dementia—developing a research protocol. Nordic Journal of Music Therapy 18(2): 103132.
  • Suzuki M, Kanamori M, Watanabe M, et al. 2004. Behavioral and endocrinological evaluation of music therapy for elderly patients with dementia. Nurs Health Sci 6: 1118.
  • Svansdottir H, Snaedal J. 2006. Music therapy in moderate and severe dementia of Alzheimer's type: a case–control study. Int Psychogeriatr 18(4): 613621.
  • Takahashi T, Matsushita H. 2006. Long-term effects of music therapy on elderly with moderate/severe dementia. J Music Ther 43: 317333.
  • Trevarthen C. 1999. Musicality and the intrinsic motive pulse: evidence from human psychobiology and infant communication. Musicae Scientiae, special issues 1999–2000: 155199.
  • Trevarthen C, Malloch S. 2008. Communicative Musicality. Exploring the Bases of Human Companionship. Oxford University Press: Oxford.
  • Vink A, Bruinsma M, Scholten R. 2003, 2011. Music therapy for people with dementia. Cochrane Database Syst Rev 2003(4). Art. No.: CD003477. DOI:10.1002/14651858.CD003477.pub2.
  • Woods RT. 2001. Discovering the person with Alzheimer's disease: cognitive, emotional and behavioural aspects. Aging Ment Health 5(S1): 716.