A systematic review of active group-based dance, singing, music therapy and theatrical interventions for quality of life, functional communication, speech, motor function and cognitive status in people with Parkinson’s disease

Background Parkinson’s disease (PD) is a common neurodegenerative condition associated with a wide range of motor and non-motor symptoms. There has been increasing interest in the potential benefit of performing arts as a therapeutic medium in PD. While there have been previous reviews, none have considered all performing arts modalities and most have focused on dance. This systematic review examined the potential benefit of all active group-based performing arts interventions for quality of life, functional communication, speech, motor function and cognitive status. Methods Searches were conducted in February 2020 on five scholarly databases. Supplementary searches were conducted. Included studies were quantitative in design, and assessed the potential benefit of any active group-based performing arts intervention for quality of life, functional communication, speech, motor function or cognitive status in people with PD. Full text papers were eligible for inclusion, as were conference abstracts since January 2018. Screening, data extraction, narrative synthesis and quality assessment were conducted independently by two reviewers. Quality assessment used the SURE checklists. Results Fifty-six studies were eligible for inclusion in this systematic review, reported in 67 publications. Published from 1989 to 2020, these studies included a total of 1531 people with PD from 12 countries, and covered four broad performing arts modalities: dance, singing, music therapy and theatre. Dance remains the most commonly studied performing arts modality for PD (38 studies), while there were 12 studies on singing interventions, four on music therapy, and only two on theatrical interventions. There was evidence for a beneficial effect of all four performing arts modalities on at least some outcome domains. Conclusions This is the first systematic review to assess the potential benefit of all active group-based performing arts interventions in PD. The evidence suggests that performing arts may be a useful therapeutic medium in PD. However, a substantial limitation of the evidence base is that no studies compared interventions from different performing arts modalities. Moreover, not all performing arts modalities were assessed for all outcome domains. Therefore it is not currently possible to determine which performing arts modalities are most beneficial for which specific outcomes.


McGill, 2019
Usual careasked not to take dance classes during the study.

McKee, 2013
Education: 20 group sessions of 90 minutes over 12 weeks. Seminars on diverse health-related topics to encourage interaction and socialising. Michels, 2018 a,b Support group: 1 group session of 60 minutes per week for 10 weeks. Traditional talking therapy support group facilitated by a professional counsellor. Patel, 2018 Education: 30 hours of group sessions over 12 weeks. Socially supportive classes addressing health and wellness topics relevant to older adults with PD. Rawson, 2019 Treadmill: 2 group sessions of 60 minutes per week for 12 weeks. Stretching: 2 group sessions of 60 minutes per week for 12 weeks.

Romenets, 2015
Self-directed exercise: a wait list control group that was additionally given a booklet about exercise in PD produced by the Parkinson Society of Canada. Shanahan, 2017 Usual care. Solla, 2019 Usual care. Ventura, 2016 Usual care. Volpe, 2013 Routine physiotherapy: individual sessions of 80 minutes covering movement, stretching, strength training, balance training, postural reeducation and gait training. Participants had an average of 21 sessions over 6 months. Westbrook, 1989 Exercise group: Structured routine of exercises including rowing movements, windmill movements of the arms, and neck exercises. The exercise classes lasted for 6 weeks. The session duration and frequency are not stated. Zafar, 2017 Usual care. Music therapy Pacchetti, 2000 Physiotherapy: weekly group sessions of 90 minutes for 13 weeks. Pantelyat, 2016 Usual care. Pohl, 2013 Usual care. Spina, 2016 Usual care. Singing Matthews, 2018 Music appreciation: watching and discussing music videos in a group once a week for 9 weekssession duration not stated. Tamplin, 2019Tamplin, ,2018 Weekly control: a weekly session of painting, dancing or tai chi. Monthly control: a monthly peer support group. Theatre Mirabella, 2017 Physiotherapy: group sessions of 1.5 hours 2 days a week for 15 months. Modugno, 2010 Physiotherapy: individual sessions of 2-3 hours 3 days a week for 3 years.

Appendix 4. Results of included studies
First author, year Results Dance Allen, 2017;McKay, 2016 There was evidence that adapted tango significantly improved motor function, including through physiological assessment. Batson, 2010 There was evidence that modern dance significantly improved balance, while the difference in TUG fell short of statistical significance. Batson, 2014 There was evidence that improvisational dance significantly improved balance (although this fell slightly short of clinical significance), while the difference in TUG fell short of statistical significance. Bearss, 2017 There was evidence that Dance for Parkinson's Disease significantly improved motor function, although this was not found for quality of life. Blandy, 2015 There was evidence that Argentine tango improved quality of life, although statistical significance was not reached.  There was evidence that a Dance for Parkinson's programme improved quality of life, although statistical significance was not reached. De Natale, 2017 There was evidence that Argentine tango was significantly more effective for motor and cognitive functions than traditional rehabilitation exercises. Duncan, 2014 There was evidence that Argentine tango was significantly more effective for motor function than usual care, and that these gains were sustained for two years while the usual control group deteriorated. Duncan, 2012;Foster, 2013 There was evidence that Argentine tango was significantly more effective for motor function than usual care. Hackney, 2007 a,b There was evidence that Argentine tango was significantly more effective for motor function than traditional exercises. Hackney, 2009 a,b,c There was evidence that tango but not American Ballroom dancing significantly improved health-related quality of life. Both tango and American Ballroom dancing significantly improved motor function versus no intervention, but the effect was stronger for tango. Hackney, 2010 There was evidence that tango significantly improved motor function (gait and balance) and that this did not differ significantly between partnered and non-partnered conditions.  There was evidence that following rather than leading tango was significantly more beneficial overall for motor function, cognition and quality of life, although leading was more effective for motor fluctuations. Hashimoto, 2015 There was evidence that PD-specific dance was significantly more effective than PD exercise or usual care in improving motor and cognitive symptoms. Heiberger, 2011 There was evidence that Dance for Parkinson's Disease significantly improved motor function (with the strongest effect being on rigidity). A significant impact on quality of life was also found, particularly relating to recreation, socialising and social impact. Hulbert, 2017;Kunkel, 2017 There was evidence that partnered dance significantly improved motor function (though not all on measures), although this effect was not found for quality of life. Kalyani, 2019 There was evidence that Dance for Parkinson's disease significantly improved cognition and quality of life compared to usual care. Koch, 2016 There was evidence that tango significantly improved quality of life, assessed by measures of well-being and body self-efficacy. Lee, 2018 There was evidence that Turo dance significantly improved quality of life compared to a waiting list control. There was some evidence for a significant benefit on motor function, being found on UPDRS-motor but not a balance assessment.

McGill, 2019
There was no evidence that ballet significantly improved motor function, considering gait and balance confidence, compared to usual care.

McKee, 2013
There was evidence that adapted tango was significantly more effective than education in improving motor function and cognition.

McNeely, 2015
There was evidence that tango significantly improved motor function, while motor symptoms deteriorated in the Dance for Parkinson's Disease group. Gait analysis variables did not however change significantly in either group. Cognitive status was assessed but results post-intervention were not reported. There was no significant effect of either dance intervention on quality of life.

McRae, 2018
There was evidence that Dance for Parkinson's Disease improves quality of life, including self-efficacy. Mediation analysis showed that one way in which higher levels of functional mobility influence overall quality of life is through enhanced self-efficacy. Marchant, 2010 There was evidence that contact improvisation dance improved motor function. Quality of life was measured but results were not reported. Michels, 2018 a,b There was evidence that a customised PD dance intervention improved motor function and quality of life, but this was not shown for cognition. The study was not designed to assess whether differences versus a support group control were significant. Patel, 2018 There was evidence that adapted tango was significantly more effective than an educational intervention for motor function, cognitive function and quality of life. Prewitt, 2017 There was evidence that a Let's Dance! programme significantly improved quality of life (self-efficacy and ADL measures) and some evidence of a significant benefit for cognitive function, although this was not found for all measures. Rawson, 2019 There was no evidence that tango significantly improved motor function or quality of life. Rocha, 2018 There was evidence that Argentine tango significantly improved mobility, balance and motor disability, while mixed-genre dance significantly improved freezing of gait. There was a trend to improved quality of life in Argentine tango participants (but not mixed-genre dance participants), although this did not reach statistical significance.

Romenets, 2015
There was some evidence that Argentine tango was significantly more effective than self-directed exercise for motor function, although this effect was not found for the primary outcome measure UPDRSmotor. There was no evidence of a significant effect on quality of life. Tango participants displayed greater improvement in cognition, although statistical significance was not reached. Shanahan, 2017 There was no evidence that Irish set dancing significantly improved motor function. However, Irish set dancing improved quality of life more than usual care, although statistical significance was not reached. Shanahan, 2015 There was evidence that Irish set dancing significantly improved quality of life. There was also an improvement in motor function, although statistical significance was not reached. Solla, 2019 Ballu Sardu offered significantly greater benefits for motor and cognitive function than usual care. Ventura, 2016 There was evidence that Dance for Parkinson's Disease was more than usual care for motor function and cognition, although effects were not consistent across measures. Large effect sizes were found for measures of cognitive switching, attention, gait speed and falls efficacy. Evidence of a significant benefit on quality of life was also found. Volpe, 2013 There was evidence that Irish set dancing was significantly more effectively than physiotherapy exercise in improving motor function. For quality of life, both groups improved, but there was no significant difference. Westbrook, 1989 There was evidence that dance/movement therapy was significantly more effective than exercise in improving movement initiation. Westheimer, 2015 There was evidence that Dance for Parkinson's Disease significantly improved motor function, but this effect was not found for quality of life. Zafar, 2017 There was evidence that adapted tango was significantly more effective than usual care in improving quality of life outcomes related to participation and autonomy, including social life, autonomy indoors and family role subscales. Music therapy Pacchetti, 2000 There was evidence that instrument-based music therapy was significantly more effective than physiotherapy for motor function and quality of life. Pantelyat, 2016 There was evidence that the drum circle intervention improved quality of life significantly more than usual care. There was some evidence that the drum circle intervention improved motor function significant more than usual care. There was no evidence for a significant beneficial effect of the drum circle intervention on cognition. Pohl, 2013 There was evidence that the Ronnie Gardiner Rhythm and Music Method significantly improved motor function, cognition and quality of life, which did not improve significantly in the usual care control group. Spina, 2016 There was evidence that active music therapy significantly improved cognition and quality of life significantly more than usual care, although this effect was not found for motor function. Singing Azekawa, 2018 There was evidence that singing improved phonatory, intelligibility and fluency, although statistical significance was not consistently reached. Di Benedetto, 2009 There was some evidence that singing significantly improved phonation, although this was not found for all phonatory measures. Elefant, 2012a,b There was some evidence that singing significantly improved functional communication (including facial expression), although this was not found for all measures. There was no evidence of a significant improvement in spoken fluency, intensity or phonatory measures.

Evans, 2012
There was evidence that singing significantly improved phonation and intensity, although this evidence was not found for intelligibility or quality of life. Higgins, 2019 There was evidence that singing significantly improved intelligibility and articulation (vowel space area). Irons, 2020Irons, ,2019 There was evidence that singing improved quality of life, with statistical significance being reached for emotional well-being, cognition and communication quality of life subscales. An effect on the social support subscale was found, but it was moderated by country with the effect being found only in South Korean and not Australian or British participants. Matthews, 2018 There was evidence that singing significantly improved phonatory, cognition and quality of life measures. Shih, 2012 There was no evidence that singing significantly improved phonatory, intensity or functional communication measures. Stegemöller, 2017a,b There was evidence that singing significantly improved motor function, quality of life, and voice-related quality of life. There was some evidence that singing significantly improved phonatory measures, although this was not found for all measures. There was no evidence of a significant benefit on swallow-related quality of life. Tamplin, 2019Tamplin, ,2018 There was evidence that singing significantly improved speech intensity and voice-related quality of life, but not phonation. Weekly participants improved more than monthly participants. Tanner, 2016 There was some evidence that singing significantly improved intensity and phonation, although this was not found for all measures. Clinically significant improvements were found for intensity range in read speech and fundamental frequency variation, while the improvement in fundamental frequency in read speech was possibly clinically significant. Yinger, 2016 There was some evidence that singing significantly improved intensity, but this was not found for all measures. There was no evidence that singing significantly improved phonation.
Theatre Mirabella, 2017 There was evidence that theatre was significantly more effective than physiotherapy in improving quality of life (including emotional wellbeing). Neither the theatre nor the physiotherapy group improved significantly in terms of motor function or cognition. Modugno, 2010 There was evidence that theatre significantly improved motor function and quality of life, whereas physiotherapy did not. Yes. Solla, 2019Ventura, 2016Volpe, 2017Westbrook, 1989Zafar, 2017 Does the study address a clearly focused question/hypothesis? Were participants/ investigators blinded to group allocation? If no, was assessment of outcomes blinded?

SURE critical appraisal checklist questions
No, partly.
Were interventions (and comparisons) well described and appropriate? Noweekly and monthly controls were different activities. Weekly and monthly interventions had leaders with different backgrounds. Was ethical approval sought and received? Yes.

b. Music therapy studies
No cohort studies.
c. Singing studies SURE critical appraisal checklist questions Azekawa, 2018Di Benedetto, 2009Elefant, 2012a,b Evans, 2012Higgins, 2019 Is the study design clearly stated? Comparison conducted on follow-up scores. Time data used for TUG. a = the primary analysis was intention to treat, though there were 9 protocol violations, of which 7 occurred in the intervention arm. Comparison conducted on change scores.. a = converted from 95% confidence interval for input into meta-analysis. b = presented by authors as a positive value as represents an improvement, but is numerically a reduction in score, and needs to be entered as a negative value in meta-analysis. c = presented by authors as a negative value as represents a deterioration, but it is numerically an increased in score, and needs to be entered as a positive value in meta-analysis.
Part C. Meta-analysis forest plots