A Cross-over Pilot Study on the Effects of Classical Guitar Instruction on Motor and Non-Motor Symptoms in People with Parkinson’s Disease

Abstract

Background:

A previous randomized pre-post cross-over study with 26 participants found positive changes in motor and non-motor symptoms in people with Parkinson’s disease (PwPD) after six weeks of group classical guitar sessions but not customary and usual treatment.

Objective:

To determine if a six-week group classical guitar instruction program improved motor function, mood, and quality of life for PwPD in comparison to a six-week group exercise program in a non-randomized cross-over pilot study.

Methods:

Eighteen PwPD were enrolled and 15 completed the study. Group 1 (N = 10) received a six-week group guitar instruction program, and then a six-week group exercise program. Group 2 (N = 8) received a six-week group exercise program, and then a six-week guitar instruction program. Assessments were at baseline, six weeks, and 12 weeks. The groups were combined for analysis by two-tailed paired t-tests due to the low sample size. Assessments included the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) motor sub-section, Hoehn and Yahr scale, Parkinson’s Disease Questionnaire-39 (PDQ-39), Apathy Evaluation Scale-Self (AES-S), and Beck Depression Inventory II (BDI-II).

Results:

MDS-UPDRS mean motor scores decreased compared to pre-test scores with group guitar instruction (−5.3 points, P < .001), but not group exercise (−0.47 points, P = .85). BDI-II mean scores decreased by 2.13 (P = .08) and 1.87 points (P = .02) with group guitar instruction and group exercise, respectively. PDQ-39 mean scores decreased by 1.93 (P = .02) and 2.52 (P = .02) points with group guitar instruction and group exercise, respectively. AES-S mean scores decreased with group exercise (−2.40 points, P = .03) but not group guitar instruction (−2.4 points, P = .26).

Conclusions:

Group guitar instruction could potentially help with both motor and non-motor symptoms in PwPD. There appears to be a specific effect of group guitar instruction on MDS-UPDRS motor scores that is not due to regular meetings and general exercises. This unfunded study was registered at ClinicalTrials.gov (NCT05917704).

Introduction

Many people with Parkinson’s disease (PwPD) are hesitant to start medications given the potential for side effects, especially early on in Parkinson’s disease. Music therapy and music-based interventions are potential alternative therapies for PwPD. In a meta-analysis on the effects of music-based movement therapy, de Dreu et al. ¹ found improvements in balance, gait speed, and stride length. Pohl et al. ² found that the Ronnie Gardiner Rhythm and Music Method, which involves rhythmic stomping and clapping to the beat of a metronome, significantly improved Unified Parkinson’s Disease Rating Scale (UPDRS) motor scores, but a control intervention did not. Bukowska et al. ³ found that neurologic music therapy focusing on therapeutic instrument performance, patterned sensory enhancement, and rhythmic auditory stimulation significantly improved balance in PwPD. In a study of the effects of music on PwPD, Pacchetti et al. ⁴ incorporated a variety of music instruments into group music therapy sessions and found significant improvement in UPDRS motor scores in the music therapy group but not the control group. On the other hand, while drumming significantly improved gait, it did not significantly improve Movement Disorder Society UPDRS (MDS-UPDRS) motor scores. ⁵ Buard et al. ⁶ found altered cortical activity and connectivity after neurologic music therapy in PwPD. Further, Maggioni et al. ⁷ found listening to music partially normalized frontotemporal interhemispheric communication in PwPD. This indicates there may be some disease modification with music therapy.

A randomized cross-over study by Bastepe-Gray et al. ⁸ found that group classical guitar instruction significantly improved motor and non-motor symptoms in PwPD. However, the study used customary and usual treatment as the control condition, which would not control for the effects of hand exercises and regular group interactions. Further, participants were not allowed to practice at home potentially limiting the amount of improvement that could be achieved. The purpose of this non-randomized unmasked cross-over pilot study was to corroborate previous research by Bastepe-Gray et al. ⁸ This study also aimed to determine what effects of group guitar training were due to physical activity or social interaction versus specific effects of guitar training by including an active group control condition focused on hand movements.

Materials and Methods

Participants

Eighteen participants met the inclusion criteria and were enrolled in the study. The target sample size (N = 30) was determined based on a previous study exploring guitar instruction in PwPD by Bastepe-Gray et al. ⁸ that had 24 participants. However, only 18 participants met the inclusion criteria and were enrolled in the study. Due to course schedules, the guitar instructor was available for a restricted time limiting the window for patient enrollment. Participants were recruited from the American Parkinson Disease Association (APDA) mailing list, the Washington State Parkinson Disease Registry, and the University of Washington movement disorder clinic.

Inclusion/Exclusion Criteria

Inclusion criteria included a diagnosis of Parkinson’s disease by a Movement Disorder Specialist and an age between 50 and 90 years old. Participants with formal music training were excluded. Participants who were unable to learn to play the guitar due to physical or cognitive impairments were excluded. The inability to physically play the guitar was not formally assessed but left to participants to decide. The cognitive inability to play the guitar included participants who could not follow instructions. Those who were receiving Parkinson’s specific physical therapy during the study period were ineligible to participate. Participants who could not attend all sessions were excluded. Participants were allowed to do other physical activities and exercises.

Ethical Approval

In accordance with the declaration of Helsinki, the experimental procedures were explained to all participants and verbal informed consent was obtained prior to participation in the study. The study protocol was approved by the Institutional Review Board (IRB) at the University of Washington (Seattle, WA; IRB ID: STUDY00017748) and registered at ClinicalTrials.gov (NCT05917704). REDCap (Research Electronic Data Capture), a secure online research data collection platform hosted at the University of Washington, was used to collect and manage study data.^9,10 Interested participants were contacted and given a written summary of the purpose, time commitment, and costs. Given the limited risk in the study, verbal consent was approved by the IRB.

Study Design

This was a non-randomized unblinded cross-over pilot study. Participants either participated in weekly group classical guitar instruction sessions for six weeks, then weekly group exercise sessions or participated in group exercise sessions, then guitar instruction sessions. Due to recruitment difficulties, participants were not randomized. The sessions were in addition to medical management. The guitar instruction sessions were 1.5 hours long and the control intervention sessions were one hour long. The subjects were found to be too tired to complete 1.5-hour group exercise sessions, so these were decreased to one-hour sessions. The guitar instructor felt one hour would not be enough to significantly improve the participants’ performances. Sessions were held at the American Parkinson’s Disease Association Headquarters Northwest Chapter.

Guitar Intervention

To mitigate the effects of Parkinson’s disease symptoms, group sessions were scheduled at a time when all participants could attend in an ON state. There was some modification of the exercises to accommodate participants who had more severe symptoms. The guitar instructional curriculum was designed by an experienced and qualified guitar instructor with progression over the classes. Classical and steel string acoustic guitars were used. Classical guitars were recommended as they have nylon strings and are therefore softer and easier to play than steel string acoustic guitars. Music selections were adapted from Learning the Classical Guitar (Buglar’s Tune, Dance of Four, Dance of Six, Serenade among other exercises). ¹¹ Other selections included Eleanor Rigby, Jolene, Let It Be, and Ode To Joy. The songs and exercises were chosen based on crossing over strings, changing chords, utilizing certain fingers, and rhythm. The participants were asked to hold the guitar in an upright position to help with maintaining posture. Each exercise started slowly and sped up to help with bradykinesia. Rhythm has been found to be particularly important for coordinating movement in PwPD. Strumming chords could help with hypokinesia. The participants also sang, which could help with hypophonia. The guitar instructor had a Bachelor of Music degree and was an Adjunct Instructor of Guitar at Seattle Pacific University. He had taught guitar lessons for 25 years. For the guitar instruction sessions, the participants were given homework and asked to practice songs and exercises at home for at least 15 minutes every day. Exercises and songs included are mentioned above. While there was no formal monitoring of homework compliance, participants were asked to perform those exercises and songs the next class to determine proficiency.

Control Intervention

The control intervention sessions began with non-rhythmic vocal exercises. Then participants completed non-rhythmic hand exercises while seated. There was some strength training using elastic bands and balance exercises. Chair yoga exercises were incorporated at the end of the sessions. The exercises incorporated hand movements like what would be used with guitar instruction. They also involved group meetings to control the effect of group interaction. The control intervention was designed to mimic the movements used when playing the guitar. There were also activities requiring multi-tasking as would be necessary on the guitar. The group exercise sessions were facilitated by an American Council of Exercise certified group fitness instructor. The instructor was also certified by the International Yoga Academy.

Outcome Measures

The primary outcome was a change in ON medication MDS-UPDRS motor scores measured by an unblinded assessor at visits zero, six, and 12. Secondary outcomes included changes in the Parkinson’s disease Questionnaire-39 (PDQ-39), Beck Depression Inventory-II (BDI-II), ¹² MDS-UPDRS modified Hoehn and Yahr Scale (H&Y Scale), and Apathy Evaluation Scale-Self (AES-S) ¹³ completed by participants at visits zero, six, and 12. Levodopa equivalent daily dose (LEDD) was calculated based on the conversion tables determined by Jost et al. ¹⁴ The assessments were done in the ON medication state as the goal was to determine whether guitar instruction would be beneficial as adjunctive therapy to medications. The MDS-UPDRS motor subsection was used to assess for bradykinesia, hypokinesia, tremor, gait, postural instability, and freezing of gait. ¹⁵ The PDQ-39 was used to assess mobility, activities of daily living, stigma, emotional well-being, social support, cognition, communication, and physical discomfort. ¹⁶ An index score was calculated incorporating all the components of the PDQ-39, where higher scores indicate worse health. The modified H&Y Scale classifies PwPD into one of five stages based on symptoms ranging from one for unilateral involvement to five for wheelchair-bound or bedridden unless aided. ¹⁷

Statistical Analyses

Paired t-tests were used to compare outcome scores before and after guitar training or group exercise. All participants were combined for these analyses, regardless of whether they started with guitar instruction or group exercise.

Results

Eighteen participants enrolled in the study with 10 in group 1 and eight in group 2. However, 15 completed the study with nine in group 1 and six in group 2. Two participants dropped out due to scheduling issues. One participant withdrew from the study as she did not feel she could play the guitar. The consort diagram is shown in Figure 1. Participants were recruited from May to October of 2023. The total study duration was eight months.

Figure 1.

Consort diagram for participant screening and enrollment.

Demographics are shown in Table 1. The age range of the participants was between 56 and 78 years old. There was no difference in the ages of participants in group 1 (mean = 66.70, SD = 6.78) and group 2 (mean = 66.38, SD = 5.34; P = .91). In group 1, 70% of participants were male. In group 2, 62% of participants were male.

Table 1.

Participant characteristics.

	Group 1		Group 2		P-value
Measure	M	SD	M	SD
Age	66.70	6.78	66.38	5.34	.91
Disease duration	7.49	6.37	3.37	3.61	.10
Sex
Male	7		5		.74
Female	3		3
Modified H&Y
1-1.5	4		4		.67
2.0-2.5	6		4
LEDD	629.2	823.52	468.75	507.75	.62
MDS-UPDRS motor	11.8	4.96	11.88	6.36	.98
BDI-II	11.7	5.74	7.13	4.67	.08
AES-S	32.6	10.82	28.5	3.51	.28
PDQ-39	17.0	8.38	11.10	9.62	.19

Abbreviations: H&Y, Hoehn and Yahr; LEDD, levodopa equivalent daily dose; MDS-UPDRS, movement disorders society unified Parkinson’s disease rating scale; BDI-II, Beck depression inventory II; AES-S, apathy evaluation scale-self; PDQ-39, Parkinson’s disease questionnaire-39.

Baseline modified H&Y scores showed that in group 1, 40% were stage one to 1.5 and 60% were stage two to 2.5. Baseline modified H&Y testing showed that in group 2, 50% were stage one to 1.5 and 50% were stage two to 2.5. All participants were Caucasian (non-Hispanic), except for one participant in group 1 and two participants in group 2 who were Asian. Ninety-four percent were right-handed. Baseline LEDD was 629.2 for group 1 and 468.75 for group 2. All participants were on a stable medication regimen or no medications prior to starting the study. The Parkinson’s disease medications patients were taking included amantadine, entacapone, selegiline, Stalevo, carbidopa/levodopa IR, carbidopa/levodopa ER, rasagiline, pramipexole, and Rytary. There were no significant changes in LEDD with the guitar or control interventions over the course of the study.

Baseline MDS-UPDRS motor scores ranged from five to 22 out of 132, with a mean of 11.83. Dyskinesias were observed during the baseline MDS-UPDRS motor exam in 22.22% of participants; however, these did not interfere with the ratings of participants.

Groups 1 and 2 were combined to compare scores prior to guitar training to those immediately following the completion of guitar lessons. For each motor and non-motor assessment, a P-value less than .05 was considered significant.

After 6 weeks of guitar instruction, the MDS-UPDRS motor scores significantly decreased by a mean of 5.3 points indicating improved symptoms (P < .001, CI [2.89, 7.78], two-tailed t-test) with a large effect size (Cohen’s d = 1.31; Figure 2a and Table 2). As shown in Figure 2a, there was a change from a mean MDS-UPDRS score of 12.07 before guitar instruction to a mean of 6.73 after the guitar instruction, which is a 44% improvement. In contrast, there was negligible difference after the control intervention (Figure 2a Table 2).

Figure 2.

Combined analysis by intervention immediately after intervention compared to before the intervention for MDS-UPDRS motor scores (a), BDI-II scores (b), AES-S scores (c), and PDQ-39 index scores (d).

***P < .001. *P < .05.

Table 2.

Combined group analysis.

	Before		After		P-value	Effect size
Measure	M	SD	M	SD
Group Guitar instruction
MDS-UPDRS motor	12.07	4.30	6.73	3.86	<.001	1.31
BDI-II	8.47	5.46	6.33	5.15	.08	0.40
AES-S	31.27	8.26	29.13	5.95	.26	0.30
PDQ-39	11.36	7.87	9.43	6.50	.02	0.27
Group exercise
MDS-UPDRS motor	9.8	5.16	9.33	5.97	.85	0.08
BDI-II	7.2	4.75	5.33	3.72	.02	0.44
AES	30.13	5.34	27.73	5.52	.03	0.44
PDQ-39	9.82	6.12	7.30	4.35	.02	0.47

MDS-UPDRS, movement disorders society unified Parkinson’s disease rating scale; BDI-II, Beck depression inventory II; AES-S, apathy evaluation scale-self; PDQ-39, Parkinson’s disease questionnaire-39.

The decrease in depression scores trended toward significance with a mean change of 2.13 points on the BDI-II after the completion of guitar instruction indicating less depression (P = .08, CI [−0.27, 4.54], two-tailed t-test) with a small effect size (Cohen’s d = 0.40; Figure 2b and Table 2). Also, there was a significant decrease of a mean of 1.87 points in BDI-II scores after the completion of the control intervention (P = .018, CI [0.37, 3.56], two-tailed t-test) with a small effect size (Cohen’s d = 0.44; Figure 2b and Table 2).

With guitar instruction, apathy decreased by a mean of 2.13 points on the AES-S, which did not reach significance, with a small effect size (Cohen’s d = 0.30; Figure 2c and Table 2). On the other hand, control group AES-S scores significantly decreased by a mean of 2.40 (P = .03, CI [0.26, 4.54], two-tailed t-test) with a small effect size (Cohen’s d = 0.44; Figure 2c and Table 2).

Quality of life, as measured by the PDQ-39, improved significantly for both groups. Following guitar instruction, PDQ-39 scores decreased by a mean of 1.93 (P = .02, CI [0.29, 3.58], two-tailed t-test) with a small effect size (Cohen’s d = 0.27; Figure 2d and Table 2). Following the group exercise program, PDQ-39 scores decreased by 2.52 points (P = .017, CI [0.53, 4.52], two-tailed t-test) with a small effect size (Cohen’s d = 0.47; Figure 2d and Table 2).

Discussion

In this small pilot study, our results suggest that group guitar instruction could have a positive impact on motor symptoms and quality of life. In contrast, group exercise did not improve motor symptoms but did lead to improvements in quality-of-life scores and reductions in depression and apathy scores. The control intervention involved a combination of voice exercises, resistance exercises, chair exercises, calisthenics, balance exercises, and chair yoga. Bastepe-Gray et al. ⁸ looked at the effects of group guitar sessions compared to customary and usual treatment on motor and non-motor Parkinson’s disease symptoms. They did not allow participants to practice outside of the sessions. In contrast, the current study compared group guitar instruction sessions to a group exercise control intervention. Further, participants were asked to practice 15 minutes a day outside of the sessions. The results found by Bastepe-Gray et al. ⁸ of significantly improved MDS-UPDRS scores and PDQ-39 scores were replicated in this study. However, the improvement in AES-S and BDI-II scores was not replicated.

Like Bastepe-Gray et al. ⁸ there was a significant improvement in MDS-UPDRS motor scores in the current study. As there was not a significant improvement in the control group exercise intervention, the improvement is likely secondary to guitar instruction rather than general exercise effects from moving their hands. It is important to distinguish between the effects of physical activity and specific guitar instruction as guitar instruction may be particularly beneficial for PwPD. The guitar intervention likely improved MDS-UPDRS motor scores due to exercises that improved hypokinesia, bradykinesia, voice volume, and gait. Exercises, such as strumming chords could help with hypokinesia. Speeding up exercises over time could help with bradykinesia. Voice volume could be improved through singing. The change of 5.33 points in the MDS-UPDRS scores represented a moderate clinically important difference. ¹⁸ Maggioni et al. ⁷ found that listening to music partially normalized frontotemporal interhemispheric communication in PwPD, which could be part of the mechanism of improvement in MDS-UPDRS scores. Additionally, Kanduri et al. ¹⁹ found that there are several genes involved in dopamine neuronal homeostasis upregulated after music performance in professional musicians, which could also be involved in the mechanism.

Unlike the study by Bastepe-Gray et al. ⁸ there was no significant improvement in depression or apathy in PwPD with guitar instruction though there was a trend toward significance for a decrease in BDI-II scores. However, there was significant improvement in non-motor symptoms with the control intervention. This indicates that the trend toward improvement in depression scores from the group guitar sessions may have been partly due to interacting with other participants every week. It is important to distinguish between the effects of social interaction versus guitar training on mood and quality of life as the benefit in these areas from group guitar training sessions may be due to the group setting rather than the guitar instruction specifically. For the BDI-II, neither intervention reached a minimal clinically important difference expected for an anti-depressant or a change of greater than or equal to three. ²⁰ There have not been studies of clinically relevant changes in apathy for comparison using the AES-S, so it is difficult to determine the clinical relevance of minor changes resulting from both interventions. For the PDQ-39, neither intervention reached the minimal clinically important difference of a decrease of 4.72 points. ²¹ The significant decrease in depression and apathy scores with the control intervention but not the guitar intervention may be because many of the participants appeared to enjoy the exercise activities more than the guitar intervention. There was also more variety of activities in the control intervention.

As with the study by Bastepe-Gray et al. ⁸ a key limitation was the small sample size. The target sample size was 30. However, due to recruitment difficulties, we did not reach the targeted sample. As the sample was low, participants were not randomized to the intervention order. With a larger sample size, decreases in apathy and depression might achieve statistical significance as they did in the study by Bastepe-Gray et al. ⁸ Also, a larger sample size would allow for the randomization and comparison of the order of intervention. In this study, the sample size required that groups be combined for analyses. Improving recruitment would also allow for randomization in future studies.

The short timeframe of the guitar instruction meant there was likely not as much of an effect size as would be seen with a longer study. Consistent with this, the effect size on depression, apathy and quality of life measures was small. This may have been partly due to selecting participants who had low baseline scores on both measures making it more difficult to see improvement. Many of the participants felt like they would benefit from more sessions.

There could be some bias with self-reported measures. Future studies could incorporate more objective measures of non-motor symptoms, such as the AES-C, which involves the clinician rating the participants’ level of apathy. Also, future studies could incorporate caregiver reports on the participants’ apathy.

A future study could include participants with deep brain stimulation devices that have the capacity to measure beta activity, which correlates with bradykinesia and rigidity. Changes in this activity could be studied in control versus guitar intervention groups.

There could be some selection bias in the study. One participant who withdrew had much more severe symptoms than other participants and would likely have not benefited as much due to difficulty with the exercises. However, this was not the case for the two other participants who withdrew. While the MDS-UPDRS rater was not masked to the treatments, which could lead to some measurement bias for motor symptoms, the non-motor symptom scores on the BDI-II, AES-S, and PDQ-39 were filled out by the participants, so would not be affected by this. Reporting bias is not an issue as there was no subgroup analysis in the study.

There was a possible dosage effect on the results that cannot be ignored as the group guitar instruction sessions were 30 minutes longer than the group exercise sessions. However, some of this additional time was used for setup, for example, tuning guitars and individual assistance due to mobility issues.

While this study had a small sample size, there were significant improvements in motor symptoms and quality of life with guitar instruction. The guitar is portable, easily accessible, and can be purchased along with accessories for around $50-$100. Guitar instruction could be more broadly implemented in community-based settings. The current study shows interest in therapeutic guitar training in the Parkinson’s disease community and the feasibility of implementing a structured curriculum in conjunction with an experienced guitar instructor. Also found by Bastepe-Gray et al. ⁸ many participants continued playing guitar after completion of the study indicating it could be incorporated into long-term treatment plans for PwPD.

Conclusions

This study corroborated a study by Bastepe-Gray et al. ⁸ on the effect of guitar-based intervention in PwPD and further showed that the effects of guitar training are likely specific and not due to general hand exercises. While this study only involved weekly sessions for six weeks, participants were given homework and asked to practice for at least 15 minutes every day. This is likely why there were still significant effects on motor symptoms and quality of life. The inclusion of a control intervention arm helped differentiate the effects of guitar versus hand exercises and the social effects of group participation. Given the significant improvement in MDS-UPDRS scores with guitar instruction compared to the control group and the significant improvement in non-motor symptoms in both groups, alternative therapies may be valuable as an adjunct to medication in the treatment of PwPD motor and non-motor symptoms. Guitar instruction could be a viable therapy, particularly for PwPD who are in the early stages of the disease.