June 2005

Singing Soothes: Music Concerts for the Management of
Agitation in Older Adults with Dementia
Kathleen A. Clarkson, BSc; Keri-Leigh Cassidy, MD; Gail A. Eskes, PhD ABSTRACT
Background
Nonpharmacological strategies for managing the behavioural and psychological symptoms of dementia are important adjuncts
to medications, and evidence to support the use of nonpharmacological options is needed. We examined the effects of music
concerts provided by volunteers on agitation in a group of demented older adults in their nursing home environment.
Methods
The study included 3 weeks of baseline, 3 weeks of daily concerts, and 3 weeks of postintervention follow-up. Agitation
among the subjects was scored using the Cohen-Mansfield Agitation Inventory during the concerts as well as weekly, during
each phase of the study.
Results
Seventeen subjects were included in the analysis. There was a near-zero incidence of agitation behaviours during the con-
certs. A decrease in the weekly physical nonaggressive agitation scores of the moderately demented subjects (Mini-Mental
State Examination score ≥15) after the introduction of music was also seen. In addition, subjects showed elevated nonag-
gressive verbal agitation through the music phase.
Conclusions
Concerts provided by untrained volunteers are a resource-friendly means of managing agitation in demented older adults.
Regular music programs might also help to trigger increased communicativeness and are a positive experience for caregiv-
ing staff and volunteers alike.
Key words: agitation, BPSD, dementia, depression, music, nonpharmacological therapy
Background
the efficacy of music for BPSD has been questioned due to Dementia is common, with an estimated prevalence of 8% in small sample sizes, short intervention periods, a limited range Canadians over 65 years of age.1 Nonpharmacological treat- of outcome measures, and inconsistent results.13–17 Most ment of the behavioural and psychological symptoms of research has been based on highly structured music therapy dementia (BPSD) is of interest, given the costs and adverse side sessions or personalized, individual music interventions.
effects of medications in older adults and especially since black These intensive approaches might not be practical for wide- box warnings were issued for the atypical antipsychotics.2,3 spread application as they demand financial, staff, and time Music is one intervention showing promise as a nonpharma- resources that might be beyond the capacities of most institu- cological alternative for managing agitation in dementia4–10 and possibly for improving mood in nondemented psychiatric Few studies have investigated music interventions taking patients.11,12 However, the quality of the existing literature on place within the context of the usual nursing home clinical set- Kathleen A. Clarkson, BSc: Dalhousie University Medical School, Halifax, Nova Antonescul, Secretary for Seniors Mental Health Team, 7th floor Abbie J. Lane Scotia; Keri-Leigh Cassidy, MD: Department of Psychiatry, Dalhousie Universi- Building, 5909 Veteran’s Memorial Lane, Halifax, NS B3H 2E2; Tel: 902-425- ty, Halifax, Nova Scotia; Gail A. Eskes, PhD: Departments of Psychiatry and Psy- 5195; Fax: 902-473-1422; E-mail: kclarkson@dal.ca chology, Dalhousie University, Halifax, Nova Scotia Conflict of interest: None declaredCan J Geriatr 2007;10:XX–XX.
Address for correspondence: Kathleen A. Clarkson, BSc, c/o Ms. Debbie C a n a d i a n J o u r n a l o f G e r i a t r i c s ting.6,8 The effectiveness of group music in activities of daily living. A geriatric cally evaluated. In the present study, we establish the degree of cognitive impair- rated by frequency from “never” (1) to reliability24 for assessment of institution- alized older adults. Categories of agita- ject’s scores. In addition to the weekly ately following lunch, 5 days a week, for behaviours were rated as “present” or ing to the Diagnostic and Statistical Manu- with pianist, consisting primarily of uni- “absent” during each concert hour.
al of Mental Disorders, Fourth Revision,18 by p.r.n. use of sedative, antipsychotic, only exclusion criterion was deafness.
contacted to explain the nature of thestudy and to seek substitute consent for Table 1. Estimated Psychotropic Drug Equivalencies Used to Calculate Total Sedatives (mg)
Antipsychotics (mg)
Antidepressants (mg)
Chart review provided demographicdata for the subjects as well as lists and time of enrolment. Regular psychotrop-ic medications were calculated by (antipsychotics), and paroxetine (antide-pressants) (Table 1). *Medication standard in each of three drug categories used to calculate equivalent doses of all other relevantpsychotropic medications in each category.
C a n a d i a n J o u r n a l o f G e r i a t r i c s M u s i c f o r t h e M a n a g e m e n t o f A g i t a t i o n i n D e m e n t i a random.25 These subjects’ MMSE scores were 8, 16, 17, and administered a similar questionnaire to the volunteer musi- 26 (i.e., three in the “moderate” and one in the “severe” demen- cians, allowing them to discuss their experiences during the tia range). The Actiwatch (model AW16, Mini-Mitter Co, Inc., Sunriver, OH) “watches” were placed on the nondominantwrist from 9:00 a.m. Monday to 4:00 p.m. Friday of weeks 3 (baseline), 6 (music), and 9 (postintervention follow-up) and Subjects’ weekly scores for the CMAI subscales and the CSDD removed only during patient bathing. The devices gathered were averaged over each of the three study phases, and the nondominant limb motion data at 1-minute intervals through- means for each subscale in each phase were first analyzed by out the 5-day period, including during sleep. For each 5-day a mixed two-by-three analysis of variance (ANOVA; within- phase of actigraph observation, mean motion counts were cal- subjects factor = study phase; between-subjects factor = culated for daytime (8:01 a.m.–8:00 p.m.) and nighttime (8:01 dementia severity group, moderate vs. severe). When these analyses indicated that the interaction effect of phase-by-groupwas significant or marginally significant (see details below), we conducted one-way repeated-measures ANOVA to exam- Weekly scores for each of the subjects on the Cornell Scale for ine the effects of phase separately for the two severity groups.
Depression in Dementia (CSDD) were completed by either A nonsignificant phase-by-group interaction was followed by or both of the same two primary nurses. The CSDD is an a one-way repeated-measures ANOVA of the two severity inventory of 19 symptoms associated with depression in groups combined. A significant one-way ANOVA was fol- patients who have dementia, and has established interrater lowed by post hoc comparisons between phases using paired- sample two-tailed t-tests. All data were analyzed using SPSSfor Windows, version 11.5.1. Postintervention Follow-UpData collection on agitation and depression continued throughout the follow-up phase. At the end of the study peri- od, a questionnaire was circulated to the care staff on the unit, Twenty-two subjects began the study; 17 completed the study evaluating their personal reactions to the intervention and per- and were included in the analyses. Completion was defined ceptions of the music’s influence on the residents. A researcher as attending a minimum of two of five concerts per week dur- Table 2. Subject Demographic and Medication Use Profiles Prior to the Baseline Phase* Moderate Dementia (MMSE ≥15) Severe Dementia (MMSE <15)
Mean daily sedative dosage (lorazepam equivalents [mg]) Mean daily antipsychotic dosage (haloperidol equivalents [mg]) Mean daily antidepressant dosage (paroxetine equivalents [mg]) *Presented as means and SDs (in parentheses).
†On the Functional Assessment Staging (FAST), a 6c corresponds to an inability to handle the mechanics of toileting.
‡Significant statistical difference between the two groups’ means (p < .05).
C a n a d i a n J o u r n a l o f G e r i a t r i c s ing the intervention phase. Two subjects failed to meet this cri- Figure 1. Percent of Subjects (n = 17) in Attendance at terion (e.g., due to having another appointment or lack of Each Concert Through the 3-week Music Phase of the assent on a given day), and three subjects died before the music phase due to complications of chronic medical illness.
Table 2 shows the profiles of the subjects according to dementia severity prior to the baseline phase. The more severe-ly demented subjects also had worse clock-drawing and FAST scores. There were no statistically significant differences in the groups’ demographic indices or medication usage.
Figure 1 shows the patterns of daily concert attendance through the 3-week music phase. The total number of concerts attended over the intervention period ranged from 8 to 15.
During the 1-hour concerts, agitation levels were extremely low. Only two and three subjects of an average of 16.6 subjects attending each concert in the first 2 music weeks showed any incidence of CMAI agitation behaviours over the 10 concerts.
Likewise, only two subjects of a mean of 15 attendees demon-strated any agitation in the third week’s concerts. When behav- iours did occur, the most common were nonaggressive Verbal nonaggression. The two-by-three ANOVA indicat-
repetitive vocalizations and mannerisms.
ed a significant effect of study phase (F[2, 30] = 4.28, p < .023) Changes in weekly physical and verbal nonaggressive and a nonsignificant interaction of phase-by-group (F[2, 30] = CMAI subscale scores were variable, depending upon the 1.06, p > .05). Follow-up t-tests to examine the phase effect (severity groups collapsed to n = 17) showed a significant rise Physical aggression. The two-by-three ANOVA indicat-
in verbal nonaggression during the music phase (baseline vs.
ed a significant effect of phase (F[2, 30] = 5.77, p < .001) and a music mean [SD] = 10.7 [5.4] vs. 13.0 [5.5], t[17] = 2.74, p < 0.015) marginally significant interaction of phase-by-group (F[2, 30] that lasted during the postintervention phase (baseline vs.
= 3.20, p < .055). One-way ANOVAs comparing the three study postintervention mean [SD] = 10.7 [5.4] vs. 12.4 [6.7], t[17] = phases for each group were significant for the moderately 2.93, p < .01; Figure 2d).
demented group only (F[2, 14] = 5.02, p < .023). Subsequent Actigraphy revealed significant variation in the response paired t-tests showed a near-significant decrease in physical to music among the four wearers (Figure 3). Due to technical aggression in the music phase compared with baseline (base- problems, only three of the wearers’ data were available in line vs. music mean [SD] = 14.0 [4.3] vs. 11.7 [1.5], t[7] = 2.23, p week 3 (baseline). There was no obvious pattern of correlation < .061) in the moderately demented group. This trend contin- between mean hourly actigraph counts and CMAI physical ued in the postintervention phase (postintervention mean [SD] agitation totals (aggressive plus nonaggressive) in either day- = 11.4 [1.1], t[7] = 2.27, p < .058; Figure 2a).
Physical nonaggression. The two-by-three ANOVA indi-
cated a significant effect of phase (F[2, 30] = 6.81, p < .004) and a marginally significant interaction of phase-by-group (F[2, 30] Over the entire study period, CSDD totals ranged from 0 to 14 = 3.19, p < .056). Follow-up one-way ANOVAs comparing out of a possible 38, corresponding to a range of “no psychi- the phases were significant for the moderately demented atric diagnosis” to “minor or probable major depressive dis- group only (F[2, 14] = 5.95, p < .013). Subsequent t-tests order.”26 No significant change was seen in the weekly CSDD revealed a significant drop in scores in physical nonaggression scores for either group (data not shown).
in the postintervention phase, relative to both the baseline(baseline vs. postintervention mean [SD] = 17.0 [7.3] vs. 13.2 [4.1], t[7] = 2.85, p < .025) and music phases (music vs. postin- On the whole, the nursing home staff reflected very positive- tervention mean [SD] = 14.5 [4.6] vs. 13.2 [4.1], t[7] = 2.46, p < ly about the perceived effects of the music intervention on the residents and on themselves (Table 3). The only Verbal aggression. The two-by-three ANOVA indicated a
unfavourable response was that the nine respondents felt the significant interaction effect of phase-by-group (F[2, 30] = 3.56, intervention marginally increased their workload, with a mean p < .04). However, subsequent one-way ANOVAs did not indi- score just below 3 (neutral) out of 5.
cate significant changes over the study period for either group The nine volunteer musicians were similarly positive in C a n a d i a n J o u r n a l o f G e r i a t r i c s M u s i c f o r t h e M a n a g e m e n t o f A g i t a t i o n i n D e m e n t i a reflecting about their experiences during the study (Table 4).
All their feedback suggested a rewarding and educational The results of this study highlight potential limitations of the measures that we currently have for behaviour and moodin dementia. We found a paradoxical increase in Discussion
verbal/nonaggressive “agitation” with the introduction of In our evaluation of live music’s effects on agitation in nursing music (see Figure 2d), despite anecdotal evidence from nurs- home–dwelling patients with dementia, we demonstrated that ing staff and families that participants were not more agitated severity of dementia was an important consideration. Follow- or difficult to manage. Rather, there were reports of positive ing the music intervention, we found a significant reduction changes in residents, such as singing to themselves after the in CMAI physical/nonaggressive behaviours in moderately concerts. There is some evidence in the literature that music demented individuals (see Figure 2b), while all subjects might improve communicativeness among older adults with showed a rise in verbal/nonaggressive agitation during the dementia. Recorded music was associated with increased ver- music phase (see Figure 2d). To our knowledge, differential balization in the 20 minutes postintervention in another study effects of music depending on dementia severity have not been of subjects with dementia.27 Informal feedback from staff and family in this study suggest that the rise in CMAI verbal “agi- Physical/nonaggressive agitation declined during the tation” might simply have reflected a positive behavioural weeks of music in the moderately demented group, but the trend did not reach statistical significance until the postinter- In addition, actigraphy and physical agitation measures vention phase. Variance found among subjects’ scores in the were poorly correlated in the few subjects for whom we had music phase might account for the lack of a statistically signif- these data. There was no clear relationship between limb icant effect. It is of interest that music continued to have bene- motion and subjective physical agitation scores (see Figure 3), fits beyond the intervention period. Previous research notes unlike what has been demonstrated in nondemented popula- music’s calming effects up to 1 hour postintervention.4,7,9 We tions.25 Similar to the difficulties faced by Brotons and Pickett- found that music interventions impacted on weekly measures Cooper,7 in the current study, p.r.n. psychotropic and analgesic Figure 2. Mean Cohen-Mansfield Agitation Inventory Subscale Scores by Study Phase Note that the maximum possible scores for each subscale differ as follows: physical/aggressive out of 77, physical nonaggressive out of 70, verbal/aggressive out of 21,and verbal/nonaggressive out of 35. Significant differences (p < .05) are noted: * = significant difference from baseline in combined groups (n = 17); + = significant dif-ference from baseline in moderate dementia group only; ++ = significant difference from music phase in moderate dementia group only.
C a n a d i a n J o u r n a l o f G e r i a t r i c s medications were administered too infrequently to be consid- Conventional BPSD measures also pose limitations to the ered a valid indicator of subject agitation. Depression in design of research in this area. Subjective rating of agitation dementia is notoriously challenging to measure, and our find- required that measures be completed by staff who observed ings did not reflect any significant benefit of music on weekly the patients closely; therefore, blinding was not possible.
observer-rated measures on mood despite anecdotal sugges- Although we used the same CMAI for the within-concert and tions that the patients’ mood had improved (see Table 3).
weekly measurements, the scale does not lend itself to statis- Although the CSDD has been validated for use in the demen- tical comparison between the two time points.
tia population,26 this conventional measure might not be sen- Given the measures’ limitations, short duration of the sitive enough to capture potential subtle changes in mood. intervention, and the study being conducted in the context of Table 3. Nursing Home Staff Feedback on the Perceived Effects of the Music Intervention* Staff Feedback Question
Mean Score (SD)†
1. How much did the residents enjoy the music? 2. How did the music affect the residents’ agitation and behaviour during concerts? 3. How much did the music impact on residents’ activity levels during concerts (e.g., trying to leave the room, moving around)? 4. How did the music affect the residents’ memory (e.g., better recall, ability to reminisce)? 5. How much impact did the music have on their mood? 6. How did the music affect the residents’ sleep? 7. How did the music affect their overall function? 9. Did the music impact on your satisfaction in your work? 10. How much did the music concerts alleviate your workload or burden? *Questions 1–7 elicited feedback on the perceived effects of the music intervention on the residents; questions 8–10 referred to effects on themselves (nursing home staff).
†All questions were rated on a scale of 1–5, with 1 being the most negative rating, 3 neutral, and 5 the most favourable rating.
Table 4. Volunteer Musician Feedback on the Personal Effects of Participating in the Study Volunteer Musician Feedback Question
Mean Score (SD)*
3. How educational was the experience in terms of: a. learning about working with older adults? b. learning about working with older adults with dementia? 4. Compared with before the concert series, how much more comfortable are you: b. interacting with older adults with dementia? 5. How likely are you to volunteer to be involved in this kind of activity again? 6. Would you recommend participating in this activity to others? *All questions were rated on a scale from of 1–5, with 1 being the most unfavourable/negative rating, 3 neutral, and 5 the most favourable/positive rating.
C a n a d i a n J o u r n a l o f G e r i a t r i c s M u s i c f o r t h e M a n a g e m e n t o f A g i t a t i o n i n D e m e n t i a usual nursing home care, the statistical- the value of subjective ratings of agita- the practicality and cost efficiency of the efits of music concerts for agitation.
ple size was used in the present investi- gation due to the constraints of a closed iours in this population. Studies of live setting, such as a music concert, are pos- shown that the benefits of live music can expand on the current literature inmusic in dementia would benefit from Figure 3. Mean Actigraph Motion Counts Per Hour* controlling for these nonspecific factorsand using larger samples sizes. In addi-tion, the use of immediate, short-, and interest. As in this study, the use of mul- measures should be included. A formalinvestigation comparing the relative fea-sibility, cost efficiency, and therapeutic Conclusions
*For each of the four actigraph wearers during representative weeks from the baseline (3), music (6), and teers, similar to the kind of interventions postintervention (9) phases. White bars represent daytime hours (8:01 a.m.–8:00 p.m.) and filled bars night- time hours (8:01 p.m.–8:00 a.m.). Total values for the individuals’ weekly Cohen-Mansfield Agitation Inventory (CMAI) physical agitation (aggressive plus nonaggressive) scores are noted above the correspon- concerts with reductions in physical agi- C a n a d i a n J o u r n a l o f G e r i a t r i c s tia and with increases in verbalization in all subjects. The Gerdner LA. Effects of individualized versus classical “relaxation” impact of the music lasted for several weeks after the interven- music on the frequency of agitation in elderly persons withAlzheimer’s disease and related disorders. Int Psychogeriatr tion, and staff and musicians alike found the experience to be a positive one that they hoped would continue. These findings 10. Suzuki M, Kanamori M, Watanabe M, et al. Behavioral and suggest that simple and low-cost interventions of music con- endocrinological evaluation of music therapy for elderly patients certs by volunteers might be an effective and practical with dementia. Nurs Health Sci 2004;6:11–8.
11. Hanser SB, Thompson LW. Effects of a music therapy strategy on approach to managing agitation in dementia, with benefits that depressed older adults. J Gerontol 1994;49:P265–9.
might extend beyond older adults with dementia to their care- 12. Hsu WC, Lai HL. Effects of music on major depression in psychi- givers and the volunteers themselves.
atric inpatients. Arch Psychiatr Nurs 2004;18:193–9.
13. Koger SM, Chapin K, Brotons M. Is music therapy an effective Acknowledgements
intervention for dementia? A meta-analytic review of literature. JMusic Ther 1999;36:2–15.
The authors gratefully acknowledge the instrumental contri- 14. Opie J, Rosewarne R, O’Connor DW. The efficacy of psychosocial butions of Dr. Ronald Stewart, MD, and of the following vol- approaches to behaviour disorders in dementia: a systematic liter- unteer musicians whose dedication and enthusiasm made the ature review. Aust N Z J Psychiatry 1999;33:789–99.
15. Lou MF. The use of music to decrease agitated behaviour of the project possible: Mary Blunden, Jonathan Brake, Allison Ged- demented elderly: the state of the science. Scand J Caring Sci des, Sarah Lea, Frederick Li, Darcy O’Brien, Ana Bela Sardinha, and Shannon Shaw. We are also grateful for the support and 16. Gräsel E, Wiltfang J, Kornhuber J. Non-drug therapies for demen- assistance of Candace Allison, RN, nursing home site coordi- tia: an overview of the current situation with regard to proof ofeffectiveness. Dement Geriatr Cogn Disord 2003;15:115–25.
nator; for the data gathering by Jacqueline Daley, RN; and for 17. Vink AC, Birks JS, Bruinsma MS, et al. Music therapy for people the help of all the nursing home staff who patiently assisted in with dementia. Cochrane Database Syst Rev 2004;3:CD003477.
this endeavour. Thanks are extended to Sarah Krieger-Frost, 18. Task Force on DSM-IV, American Psychiatric Association.
RN, for helping with the preliminary data collection and to Diagnostic and Statistical Manual of Mental Disorders: DSM-IV,4th edition. Washington, DC: American Psychiatric Association; Michael Lawrence and Dr. John Christie for statistical and tech- nical advice. Thanks to Dr. David Gardner for his assistance 19. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” A with pharmacological equivalency estimates of psychotropic practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–8.
20. Reisberg B. Functional Assessment Staging (FAST). Psychophar- Kathleen A. Clarkson’s contribution was funded by a Nova Scotia Association of Health Organizations scholarship 21. Shulman KI. Clock-drawing: is it the ideal cognitive screening in continuing care research. Gail A. Eskes is supported by a test? Int J Geriatr Psychiatry 2000;15:548–61.
Faculty of Medicine Clinical Research Scholar Award.
22. Cohen-Mansfield J, Billig N. Agitated behaviors in the elderly. I. A conceptual review. J Am Geriatr Soc 1986;34:711–21.
References
23. Finkel SI, Lyons JS, Anderson RL. Reliability and validity of the Cohen-Mansfield Agitation Inventory in institutionalized elderly.
Lindsay J, Sykes E, McDowell I, et al. More than the epidemiology Int J Geriatr Psychiatry 1992;7:487–90.
of Alzheimer’s disease: contributions of the Canadian Study of 24. Cohen-Mansfield J, Marx MS, Rosenthal AS. A description of agi- Health and Aging. Can J Psychiatry 2004;49:83–91.
tation in a nursing home. J Gerontol 1989;44:M77–84.
Opie J, Doyle C, O’Connor DW. Challenging behaviours in nurs- 25. Grap MJ, Borchers CT, Munro CL, et al. Actigraphy in the critical- ing home residents with dementia: a randomized controlled trial ly ill: correlation with activity, agitation, and sedation. Am J Crit of multidisciplinary interventions. Int J Geriatr Psychiatry 26. Alexopoulos GS, Abrams RC, Young RC, et al. Cornell Scale for Livingston G, Johnston K, Katona C, et al. Systematic review of Depression in dementia. Biol Psychiatry 1988;23:271–84.
psychological approaches to the management of neuropsychiatric 27. Sambandham M, Schirm V. Music as a nursing intervention for symptoms of dementia. Am J Psychiatry 2005;162:1996–2021.
residents with Alzheimer’s disease in long-term care. Geriatr Gerdner LA, Swanson EA. Effects of individualized music on con- fused and agitated elderly patients. Arch Psychiatr Nurs 28. Cohen-Mansfield J, Libin A. Assessment of agitation in elderly patients with dementia: correlations between informant rating Groene RW. Effectiveness of music therapy 1:1 intervention with and direct observation. Int J Geriatr Psychiatry 2004;19:881–91.
individuals having senile dementia of the Alzheimer’s type. J 29. Woods B, Spector A, Jones C, et al. Reminiscence therapy for dementia. Cochrane Database Syst Rev 2005;2:CD001120.
Goddaer J, Abraham IL. Effects of relaxing music on agitation dur- 30. Van de Winckel A, Feys H, De Weerdt W, et al. Cognitive and ing meals among nursing home residents with severe cognitive behavioural effects of music-based exercises in patients with impairment. Arch Psychiatr Nurs 1994;8:150–8.
dementia. Clin Rehabil 2004;18:253–60.
Brotons M, Pickett-Cooper P. The effects of music therapyintervention on agitation behaviors of Alzheimer’s diseasepatients. J Music Ther 1996;33:2–18.
Clark ME, Lipe AW, Bilbrey M. Use of music to decrease aggres-sive behaviors in people with dementia. J Gerontol Nurs1998;24:10–7.
C a n a d i a n J o u r n a l o f G e r i a t r i c s

Source: http://oneinfive.ca/wp-content/uploads/2011/03/Singing-Soothes-KLC.pdf