Music Intervention for Disruptive Behaviors in Long-Term Care Residents with Dementia

Disruptive behaviors, defined by Rossby et al¹ as “behaviour resulting in negative consequences for the resident, caregiver or other residents,” have several negative consequences in long-term care (LTC) residents. For example, they contribute to resident and staff stress, which may lead to staff burnout² and increased occurrence of injury secondary to falls.³ Although restraint use has been discouraged since the Omnibus Budget Reconciliation Act of 1987 (OBRA ’87),⁴ pharmacologic and physical restraints are still used to manage disruptive behaviors.^5-8 Physical restraints have been associated with increased agitation^8,9 and injuries.^10,11 Similarly, pharmacologic restraints, such as antipsychotics, mood stabilizers, and benzodiazepines, are often not well tolerated by the elderly.^12-14 Recently, off-label use of atypical antipsychotics has further been discouraged by the Food and Drug Administration after 15 out of 17 placebo-controlled trials found that these medications increase mortality when used to treat behavioral disturbances in elderly residents with dementia.¹⁵ Given the adverse effects of physical and pharmacologic restraints and OBRA restrictions, it is prudent to investigate alternate methods of managing disruptive behaviors. Music may be a relatively simple, noninvasive, and inexpensive intervention for disruptive behaviors that does not require a huge time expenditure. This article reviews the evidence for preserved musical ability in elderly LTC residents with dementia, and the effect of music intervention on disruptive behaviors.

It has been proposed that even in advanced stages of dementia, certain levels of musical perception may still be retained in the absence of hearing impairment.¹⁶ There is anecdotal evidence that music cognitive ability may be preserved in at least a subset of individuals with dementia, despite deficits in simpler cognitive tasks. Beatty et al^17,18 and Crystal et al¹⁹ described preserved musical ability in individuals with dementia who had cognitive deficits, such as loss of language function and inability to dress. In stroke patients, it has been observed that the areas of the brain involved in verbal memory are different than the areas involved in musical memory.²⁰ This suggests that patients with a loss of language function secondary to Alzheimer’s disease may have preserved musical abilities. There are several mechanisms by which music intervention may affect individuals with dementia. For example, music may evoke pleasant memories²¹ or induce relaxation.^22,23

Several studies have demonstrated that music intervention in elderly persons with dementia improves social behaviors,24 enhances language function,²⁵ and increases spatial-temporal reasoning.²⁶ There are also several studies showing that music intervention may reduce agitation and aggression among long-term care residents with dementia.

For example, Brotons and Pickett-Cooper²⁷ conducted group music therapy sessions for 20 nursing home residents with dementia, and found that they were significantly less agitated during the sessions, independent of whether or not they had a musical background. This study had a small sample size, and insufficient information was collected to determine whether or not music therapy sessions affected as-needed medication. Further, as there was no control group in this study, it is difficult to say whether the reduction in agitation was due to music intervention or to some other factor.²⁷

Tabloski and colleagues²⁸ used a quasi-experimental design where each subject served as their own control, and found that calming music significantly reduced agitation in nursing home residents. This study also used a limited sample size (20 individuals), and while scores on the Agitated Behavior Scale were found to be significantly different, it is not clear if this correlated with a clinically relevant reduction in agitation.²⁸

Some authors have found that aggressive behaviors most likely occur in the morning, when staff violated residents’ personal space in order to assist them with activities of daily living such as toileting, bathing, dressing, and eating.²⁹ Lack of cooperation may prevent residents from receiving assistance with these activities. Several authors have examined the effects of music intervention during mealtime and during bathing. Denney³⁰ found that there was a 46% decrease in agitation when quiet music was played during lunchtime. The effect was seen with verbally agitated behaviors and physically nonaggressive behaviors. Goddaer and Abraham³¹ used a repeated-measures design to expose nursing home residents to relaxing music during mealtime. They found significant reductions in total agitated behaviors, physically nonaggressive behaviors, and verbally agitated behaviors; however, there was no change in aggressive behaviors and hiding/hoarding behaviors. Ragneskog and associates³² played three different types of music (pop, Swedish tunes, and soothing music) for two-week periods during dinnertime at a nursing home. It was found that during all music periods, residents ate more food in total, possibly because staff were serving the residents more food. In addition, authors found that residents appeared less irritable and anxious, especially when soothing music was played.

To date, there are only two documented studies in the literature on the effects of music during bathing of cognitively impaired nursing home residents. Thomas et al³³ observed 14 residents during three baseline bathing episodes, three treatment (music intervention) periods, and three post-treatment bathing episodes. Using a modified form of the Cohen-Mansfield Agitation Inventory (CMAI),³⁴ investigators found significant reductions for aggressive behavior, but not for hiding/hoarding behavior or physically nonaggressive behavior during music treatment sessions.³³ Clark et al³⁵ observed 18 residents during ten bathing episodes with “preferred” music and ten bathing episodes without music, and found a reduction in aggressive behaviors during the music condition. In addition, caregivers reported more cooperation from residents. Limitations of this study include small sample sizes, lack of a blinded study design, effects of psychotropic medication not taken into consideration, and the huge intra-subject and inter-subject variability in the population.³⁵

Gerdner³⁶ used a repeated-measures, crossover design to compare the effects of classical “relaxation” music versus familiar or “individualized” music on agitated behaviors in LTC residents with dementia. It was found that both music conditions reduced agitated behaviors; however, the effect was stronger for the individualized music.³⁶ Agitation was measured using a modified version of the CMAI. This scale was originally designed to assess agitated behaviors over a two-week period, and the psychometric properties of the modified CMAI have not been established. Again, while the differences in modified CMAI scores were statistically significant, it is unclear whether or not they were clinically significant.

In a previously unpublished pilot study completed by the author, headphones were used to single-blind a randomized, controlled, preferred music intervention for disruptive behaviors in LTC residents with severe dementia. The Disruptive Behavior Scale (DBS) was chosen as a measure of the disruptive behaviors because it records not only the frequency of the behaviors, but also their severity.³⁷ A significant change in score on the DBS was determined based on caregiver comments, as-needed medication use, and physical restraint use.

There was a trend toward a reduction in disruptiveness during the no-music and music conditions, the effect being greater for the music condition in comparison to the baseline period. In 28% of the observations, during the baseline period, subjects scored below 300 (ie, low level of disruptiveness) on the DBS, as compared to 46% during the music condition. A sample of 141 (power 0.99, alpha 0.05) is needed to make these results significant. Residents who participated in the study made comments such as, “Oh, look, they are singing,” and “Can we do this again tomorrow?”

CONCLUSIONS

Many studies have shown that music intervention, particularly preferred music intervention, may reduce disruptive behaviors in elderly individuals with dementia. A review of the literature, however, shows that these previous studies lack rigorous research designs, which make it difficult to draw clear conclusions from the results.38 For example, previous studies tended not to be randomized, lacked control groups, lacked blinded study designs, and had limited sample sizes. As described earlier, it is possible to use headphones in an attempt to “blind” the music intervention in future studies.

Most music intervention studies used behavior rating scales that have not been validated, nor did they include dependent variables such as caregiver burden, as-needed medication use, and restraint use, which are factors that may have indicated if the music intervention was clinically significant. Further research is needed to clarify the role of dementia severity and the mode of music intervention.

Despite the methodological limitations of previous studies, it appears that music is a promising alternative to restraint use in the management of disruptive behaviors. It may be argued that if trained music therapists are used to provide supervision or direct music therapy, then music intervention is not necessarily a simple or an inexpensive treatment for disruptive behaviors. However, this in no way would diminish the potential value of music intervention to modify disruptive behaviors, provide cognitive stimulation, and improve the quality of life for elderly LTC residents with dementia.

Acknowledgments
The author would like to thank Christine Brownell, Gail Edwards, Winnie Kwan, Brent Peattie, and Elaine Thomas for data collection; Dr. Rollin Brant for advice regarding data analysis; and the patients and staff of the North Haven’s Unit, Bethany Care Centre, Calgary, Alberta.

The research reported in this article was supported by a grant from the Calgary Health Region Adult Research Committee, Centre for Advancement of Health, Calgary, Alberta. It was presented at the Canadian Academy of Geriatric Psychiatry Annual Meeting on November 4, 2002, in Banff, Alberta, and the Alberta Psychiatric Association Annual Meeting in 2003, in Banff, Alberta, Canada.