Music Therapy For Dementia Symptoms Treatment
Overview
The study explores the impact of individualized music listening (IML) on the behavioral and psychological symptoms of dementia (BPSD), a topic previously assessed through proxy-completed questionnaires. To address methodological limitations, the researchers employed a systematic observational approach, incorporating a validated instrument, time-based sampling, and longitudinal analytical methods.
Conducted as a randomized controlled trial, the study involved nursing home residents with dementia assigned to either an IML intervention group (IG; n=44) or a control group (CG; n=46). Trained raters observed 18 BPSD in 15 four-minute intervals before, during, and after an IML session. The comparison between the IG and CG utilized t-tests for different time points and piecewise latent curve modeling to assess BPSD trajectories over an hour.
Results indicated that BPSD occurred less frequently in the IG during the session but not before or after. Additionally, while the CG exhibited a stable trajectory of BPSD, the IG displayed a u-shaped pattern, characterized by stability before the session, a decrease during the session, and an increase after. Significant interindividual variability was noted in baseline BPSD and pre- and post-session slopes.
The findings contribute additional evidence supporting the effectiveness of IML in reducing BPSD, despite the short-lived nature of the effect. Notably, IML is highlighted as a positive intervention due to its lack of negative side effects, high acceptance, and ease of implementation. The study suggests the integration of IML into the everyday care routines for individuals with dementia.
Introduction
Dementia stands as a significant global health challenge, adversely affecting the quality of life for individuals with dementia (PwD) and their caregivers, while also imposing substantial financial burdens on society. While dementia is primarily characterized by cognitive decline, it often manifests alongside behavioral and psychological symptoms of dementia (BPSD), encompassing clusters such as paranoid ideations, aggression, hallucinations, agitation, anxieties, phobias, and affective disturbances. These symptoms significantly impact the well-being of PwD and contribute to caregiver burden, even after institutionalization.
Recognizing the limitations and adverse effects of pharmacological treatments, there is a growing interest in non-pharmacological interventions to alleviate BPSD. Music-based interventions, particularly individualized music listening (IML), have garnered attention for their potential efficacy. IML involves PwD listening to personalized music playlists based on their preferences and experiences. Several reviews suggest that IML effectively reduces BPSD, such as anxiety and agitation, by stimulating positive memories and reducing stress.
However, existing studies on IML often face methodological challenges, including small sample sizes, single-group designs, and reliance on proxy-completed questionnaires, which may introduce bias. Although eight studies have explored the immediate effects of IML through observational methods, these studies also have limitations. Most are based on small samples and lack psychometrically evaluated observational measures. Additionally, previous studies utilizing time-sampling approaches did not address the overall shape of the growth curve or interindividual differences in BPSD trajectories.
The current study aims to address these gaps by conducting a well-powered randomized controlled trial (RCT) to assess the impact of IML on BPSD in nursing home residents with advanced dementia. Trained raters systematically observed a diverse range of BPSD in short time intervals before, during, and after IML sessions. The research explores static differences between PwD receiving the intervention and those receiving standard care at various time points. To provide new insights into the latency and duration of IML effects and interindividual variability in trajectories of change, piecewise latent change growth models (PLCGM) are employed, offering a more nuanced analysis compared to traditional ANOVA methods.
Existing evidence supports IML as an effective non-pharmacological intervention for reducing BPSD. Nevertheless, the current study seeks to overcome previous methodological limitations by employing a robust RCT design and psychometrically evaluated observational measures to enhance the understanding of IML’s impact on BPSD in individuals with advanced dementia.
Methods
The study involved residents from five nursing homes in Germany diagnosed with dementia, recruited for a randomized controlled trial (RCT) registered retrospectively in the German Clinical Trials Register and ISRCTN registry. Individuals with severe hearing problems were excluded from participation. The participants were randomly assigned to either an intervention group (IG) or a control group (CG) using a computer-generated randomization list, stratified by gender. The IG engaged in individualized playlist sessions through headphones every second day for six weeks, while the CG received standard care, which could include musical activities later offered to the IG as well.
Ethical approval was obtained from the Ethics Committee of the Faculty of Social and Behavioral Sciences of the Friedrich Schiller University Jena, with participants or their legally authorized representatives providing written informed consent. The sample comprised N=90 people with dementia (PwD), and there were no significant differences between the IG and CG in terms of age, gender, or cognitive functioning.
Behavioral observations were conducted using the challenging behaviors subscale of the Dementia Coding System (DeCS), a coding system designed to systematically evaluate non-pharmacological interventions for PwD. Three psychologists and four student assistants, trained by experts in the DeCS, assessed participants’ BPSD on three occasions, once every other week during the IML intervention. The DeCS utilized a time-sampling approach, observing participants for 15 consecutive four-minute time units (TUs), summing scores for 18 different BPSD (e.g., crying, resistance, agitated behaviors).
The IG participants underwent IML sessions during TUs 6 through 10, with observations conducted for 20 minutes before, during, and after the session. Inter-rater reliability was excellent, and the average BPSD per TU was calculated across observation occasions. No significant differences were found in pre-session BPSD scores across observation occasions.
The study employed rigorous methods, including randomization, ethical approval, and behavioral observations using the DeCS, to investigate the impact of individualized music listening on BPSD in nursing home residents with dementia. The research aimed to provide a detailed understanding of BPSD before, during, and after IML sessions.
Statistical Analysis
The data preparation steps were executed in SPSS (IBM, Version 28.0), and statistical analyses were performed in R. Little’s test was employed to examine whether data were missing completely at random (MCAR). Listwise deletion and imputation using the expectation-maximization algorithm were compared to assess the potential for bias due to missing data.
The study hypothesized that individualized music listening (IML) would decrease behavioral and psychological symptoms of dementia (BPSD) compared to standard care. One-sided t-tests and the “car” package in R were utilized to evaluate whether BPSD were lower in the IML group during and after the IML session. A significance level (α) of 0.05 was set, and Cohen’s d was used as a measure of effect size.
Piecewise latent change growth models (PLCGM) and the “lavaan” package in R were employed to compare trajectories of BPSD in the IML and control groups across 15 time units (TUs). The PLCGM allowed for the definition of multiple phases of change, and two unconditional models with differing constraints on equal residual variance were compared. An intercept and three slope factors representing linear changes before, during, and after the IML session were specified in the models.
Given the non-normal distribution of the continuous outcome variable (BPSD), a maximum likelihood estimator with robust standard errors and corrected model test statistics was used. The model fit was assessed using Hu and Bentler’s two-index presentation strategy, with a comparative fit index (CFI) ≥ 0.90 and a standardized root mean square residual (SRMR) ≤ 0.08 indicating reasonable fit. The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were considered as comparative fit indices, where lower values suggest better fit. R2 values of the manifest variables were used to assess the proportion of variance in BPSD at each TU explained by the model.
Results
The study employed data from five nursing homes in Germany, utilizing a randomized controlled trial (RCT) design. Little’s MCAR test indicated no evidence that data were missing completely at random (p = 0.080). Both listwise deletion and imputation of missing data yielded comparable results, leading to reporting based on complete data. The primary hypothesis was that individualized music listening (IML) would decrease behavioral and psychological symptoms of dementia (BPSD) compared to standard care.
Analysis of BPSD before, during, and after the IML session revealed that BPSD were rarely observed in either the intervention group (IG) or control group (CG) before or after the IML session. However, BPSD were significantly lower in the IG compared to the CG during the IML session (Cohen’s d = 0.45).
BPSD trajectories across 15 time units (TUs) were visualized in Figure 2, showcasing distinct, non-linear trajectories for the IG and CG. Piecewise latent change growth models (PLCGM) were employed to model BPSD changes before, during, and after IML. The unconditional PLCGM with unconstrained equal residual variances exhibited better fit than the model with constrained equal residual variances.
According to the unconstrained model, there was a significant decrease in BPSD during IML, with substantial interindividual variation in baseline BPSD. The conditional PLCGM, including IG as a predictor, demonstrated reasonable fit and accounted for variability in BPSD. IG was associated with a reduction in BPSD during and after the IML session, indicating a positive impact of IML on BPSD. However, IG was not significantly associated with baseline BPSD or pre-session changes in BPSD.
Conclusion
This randomized controlled trial (RCT) represents the first investigation of the immediate effects of Individualized Music Listening (IML) using a psychometrically evaluated observational measure of Behavioral and Psychological Symptoms of Dementia (BPSD) in nursing home residents. The study employed the “challenging behaviors” subscale of the Dementia Coding System (DeCS) to assess a range of BPSD, such as agitation, depression, aggression, and disorientation.
The findings extend prior research by focusing on the immediate effects of IML, specifically on agitation, a frequently observed neuropsychiatric symptom. Additionally, the study provides unique insights into the latency and duration of IML effects, considering interindividual differences in how people with dementia (PwD) react to IML. Both t-test and Piecewise Latent Change Growth Model (PLCGM) analyses suggest that IML has a meaningful and short-term reduction effect on BPSD, particularly in agitation.
Compared to PwD receiving regular care, those participating in the IML intervention exhibited similar BPSD before and after the IML session, but lower BPSD during the session. The PLCGM analyses confirmed these results, revealing a u-shaped trajectory in the IML group characterized by stability before the session, a decrease during the session, and an increase after the session. This suggests a short-term impact of IML on BPSD, with post-session BPSD somewhat lower than baseline, although not statistically significant.
The study aligns with other research indicating that the effects of IML on BPSD are immediate but short-lived. Notably, emotional reactions to meaningful music seem to be evoked during listening rather than persisting afterward. The observed u-shaped change in BPSD further supports the short-term nature of the effect. The study’s findings contribute valuable insights into the potential of IML as a non-pharmacological intervention for managing BPSD in individuals with dementia.
Oncology Related Tools
Other
Latest Research
