Short report

Group therapeutic singing improves clinical motor scores in persons with Parkinson’s disease

Abstract

Background Previous reports suggest that group therapeutic singing (GTS) may have a positive effect on motor symptoms in persons with Parkinson’s disease (PD).

Objective To determine the effect of a single session of GTS on clinical motor symptoms.

Methods Clinical motor symptom assessment was completed immediately before and after 1 hour of GTS in 18 participants.

Results A significant decrease in average scores for gait and posture and tremor, but not speech and facial expression or bradykinesia was revealed.

Conclusion These results support the notion that GTS is a beneficial adjuvant therapy for persons with PD that warrants further research.

Introduction

Incorporating music into current treatment strategies for Parkinson’s disease (PD) is beneficial, yet the underlying mechanism remains a challenge to understand. Various forms of dance and drumming improve functional mobility, postural instability and walking rate, while singing improves voice, respiratory control and swallow.1–6 These results are domain specific demonstrating improvements in outcome measures that are associated with the exercise/therapy or target area. Given that increased physical activity has many positive benefits in persons with PD,7 detangling the effects of music from the effects of increased physical activity/exercise is difficult.8 Examining the effects of music on non-domain specific outcome measures may provide further direction in understanding underlying mechanisms.

Group therapeutic singing (GTS) involves sitting and singing within a group for an hour, with no overt physical activity. In previous studies, Stegemöller et al found improvements in voice, respiratory control, swallow, quality of life after 8 weeks of GTS.1 2 Participants in these studies also reported that the groups were fun and relieved stress, they felt better and were able to move better after GTS.3 These results would suggest that GTS may have an effect on multiple symptoms, including motor symptoms of PD. Examining the effects of GTS on clinical motor symptoms may differentiate the effects of singing from physical activity, providing further insight into how music alone benefits person with PD. The purpose of this pilot study was to determine the effect of aa single session of GTS on the clinical evaluation of speech and facial expression, bradykinesia, gait and posture, and tremor using the Movement Disorders Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS).9 Given the previous results of GTS, we hypothesised that only speech and facial expression will improve after 1 hour of GTS.

Methods

Participants

Eighteen participants (11 female, 7 male; age=74.9±7.2 years, 100% right handed, 100% white) diagnosed with PD by their treating neurologist were enrolled into the study. No participant demonstrated signs of severe cognitive impairment (Mini-Mental State Examination=28.4±1.8) or depression (Beck Depression Inventory=10.9±3.7) (supplementary table 1).

Participants were on the same PD medication as prescribed by their treating physician for 30 days prior to data collection (supplementary table 3). The average disease duration was 7.5±4.7 years and 53% reported that the right side was their most affected side. The average total score on the MDS-UPDRS (parts I–IV) was 73.4±13.6. All participants took their normal medication and their normal times, while the GTS session remained fixed at the same day and time. Thus, all participants were tested on medication, on average 206.2±92.4 min since taking their last dose (supplementary table 2). No participant took additional medication throughout the duration of the data collections or GTS session. Participants had been participating in GTS for an average of 2.4±1.4 years prior to the date of data collection.

Singing session

The GTS session lasted 1 hour. The singing session consisted of a greeting song (5 min), a series of vocal exercises (15 min), therapist chosen group singing (15 min), participant chosen group singing (20 min) and a closing song (5 min). The session was led by a board-certified music therapist who used piano as the primary accompanying instrument. No lyrics or music were provided to promote proper posture while singing. The vocal exercises and therapist chosen songs have been used in previous GTS studies and a more detailed description can be found in these studies.1–3

Movement Disorders Society Unified Parkinson’s Disease Rating Scale

Motor examination (part III) of the MDS-UPDRS was videorecorded immediately before and after the GTS session. Videos were coded to mask pre or post intervention order and were scored by two movement disorders neurologists that were masked to the study intervention. Since the MDS-UPDRS-III was recorded, scores for rigidity were not completed. Independent t-tests revealed no differences between the raters’ scores (presession scores: p=0.48; postsession scores: p=0.55). Thus, the average of the scores was used for the remaining analyses. The speech and facial expression score was calculated as the sum of items 3.1 and 3.2. The bradykinesia score was calculated as the sum of items 3.4 through 3.8 plus 3.14. The gait and posture score was calculated as the sum of items 3.9 through 3.13, and the tremor score was calculated as the sum of items 3.15 through 3.18.

Statistical analysis

The Shapiro-Wilk test revealed that all distributions were normal. Paired t-tests probed for differences between pre and post MDS-UPDRS-III scores for the total motor score, and each domain (speech and facial expression, bradykinesia, gait and posture, and tremor). Effect sizes using Cohen’s d were calculated. Significance was set a p<0.05. Change scores were calculated as the post score minus the pre score for descriptive results. Because participants were on medication, Pearson correlations were completed for time since medication and the change score for each MDS-UPDRS-III domain score and the total motor score. In addition, because participants had been singing for different lengths of time, Spearman correlations (years singing was not normally distributed) were completed for previous years singing and the change score for each MDS-UPDRS-III domain score and the total motor score.

Results

The mean, SE and statistical results for all comparisons are shown in table 1. Individual participant change scores for the total motor score and each domain are shown in figure 1 and supplementary table 2. For the total motor score, 67% of the participants demonstrated an improvement (ie, decrease in score=−1.75) in motor symptoms. Results indicated a trend (p=0.09, d=0.19) towards significance. However, it is important to note that the largest individual participant improvement in total motor score was −11.5 points, and the largest worsening was +3.5 points. Results were variable across participants for domain scores, as changes across each domain were different for each participant. Nonetheless, results revealed a significant improvement in average scores for gait and posture (−0.62, p=0.03, d=0.25) and tremor (−1.28, p=0.05, d=0.27). For gait and posture, 56% of the participants demonstrated an improvement in score, and for tremor, 50% of the participants demonstrated an improvement in score. Average scores worsened slightly for bradykinesia (0.11) and speech and facial expression (0.03). For bradykinesia, 50% of the participants demonstrated an improvement in score, while only 33% of the participants demonstrated an improvement in score for speech and facial expression.

Figure 1
Figure 1

The change in individual participant MDS-UPDRS-III scores from pre-GTS to post-GTS for the (A) total motor score, (B) speech and facial expression score, (C) bradykinesia score, (D) gait and posture score, and (E) tremor score. GTS, group therapeutic singing; MDS-UPDRS, Movement Disorders Society Unified Parkinson’s Disease Rating Scale.

Table 1
|
Mean, SE, range and paired t-test results for MDS-UPDRS-III scores

No significant correlations were revealed between time since medication and the change score for total MDS-UPDRS-III score (R=−0.09, p=0.73) nor each domain score (R≤0.23, p>0.36). There was no significant correlation between years singing and the change score for the total MDS-UPDRS-III score (R=−0.19, p=0.4) However, for years singing, a significant negative correlation was revealed for bradykinesia (R=−0.47, p=0.049) and a significant positive correlation was revealed for gait and posture (R=0.56, p=0.016). Those who had been singing longer showed a greater improvement in bradykinesia, while those with less experience singing showed a greater improvement in gait and posture. There were no other significant correlations between years singing and tremor (R=−0.14, p=0.6) nor speech and facial expression (R=−0.17, p=0.5).

Discussion

The purpose of this study was to determine the effect of a single session of GTS on the clinical evaluation of motor symptoms in persons with PD. Given that research has shown that GTS improves the voice in person with PD and GTS does not involve overt physical activity, we hypothesised that only speech and facial expression would improve after 1 hour of GTS, but not bradykinesia, gait and posture, and tremor. However, results were in contrary to our hypothesis. Scores for speech and facial expression, as well as bradykinesia, did not significantly change. Scores for gait and posture and tremor did significantly decrease and there was a significant association between years singing and gait and posture. Those with less experience singing showed the greatest improvement in gait and posture. The total motor score decreased, as well, and was close to significance. These results suggest that GTS may immediately engage and/or enhance other neural networks that are involved with gait, posture and tremor. Improvements in intervention specific symptoms (ie, speech, facial expression, voice, swallow, respiratory control) may take a longer period of time over several sessions before observing significant changes, as demonstrated in previous studies, or may be the result of fatigue after an hour of singing.1–3 Moreover, results do support that improvements in bradykinesia may emerge over time given the significant association between years singing and the change in bradykinesia score. Those who had been singing long showed greater improvements in bradykinesia. Taken together, this supports the need for future long term research on the effects of GTS on motor symptoms.

An interesting finding of this study was that improvements in clinical motor symptoms were observed while participants were on their optimal PD medication dosage. There were no significant associations between time since medication and change scores, suggesting that GTS may account for the improvement in motor symptoms. However, since the timing of medication was not controlled and the dosage was not collected, it is likely that not all participants were in their peak state which may explain the variable response. Nonetheless, some participants demonstrated improvements in the total motor score up to 11.5 points while on medication. This promising result supports the need for further research aimed at optimising GTS for persons with PD. Understanding the underlying neurophysiology may be the first step.

Research has shown that engaging in music increases activity within the mesolimbic system, specifically the ventral tegmental area (VTA) and nucleus accumbens (NAcc).10 11 These nuclei have connections with the basal ganglia as part of the limbic loop,12–14 which may explain why singing improved gait, posture and tremor. Efferent projections from the NAcc target the output nuclei of the basal ganglia and the VTA has reciprocal connections with the striatum.12 13 Research has traditionally suggested the functional loops with in the basal ganglia are segregated circuits.13 However, there is increasing evidence that these circuits are not segregated and play an important role in modulating motor control based on mood, memory and cognition.12 Indeed, research has shown that a unidirectional influence of the limbic loop over the motor loop in animal models.15 Thus, a possible explanation of why a single session of GTS improved gait and posture and tremor in persons with PD may be that increased activity of the VTA and NAcc modulated activity within the limbic loop of the basal ganglia in turn influencing or compensating for impaired activity within the motor loop. This proposed underlying mechanism of how GTS improves motor symptoms in PD are truly speculative based on limited research, andthere remains a need for continued research.

Limitations

While this was a preliminary study, there are limitations to consider. The criteria for diagnosis of PD is unknown due to the nature of the study site. There is the possibility that other neurodegenerative diseases may have been present in the participant sample. In addition, we did not control specifically for the timing or dose of medication, but rather maintained the same day and time for the GTS intervention. For both of these limitations, the study design reflects the typical delivery of the intervention in which GTS is offered at a particular time regardless of each individual’s medication and the participants may have other diagnoses other than PD that are unknown at the time of the intervention. Finally, given the nature of the intervention and study design, there is the potential of a placebo effect and study bias. Further long-term research including an appropriate control is needed.

Conclusion

The results of this study provide the preliminary behavioural evidence suggesting that GTS improves motor symptoms beyond the typical pharmacological effect for some persons with PD. Any improvement in symptoms is worthy of continued research to further the understanding and implementation of music and singing in the treatment of persons with PD.