Potential Benefit of Singing for People with Parkinson’s Disease: A Systematic Review

INTRODUCTION

In an increasingly ageing global population [1, 2], age-associated conditions such as Parkinson’s disease (PD) are a major public health challenge of the future. Between 7 and 10 million people worldwide are estimated to have PD [3] and it is now recognisedto be a multifaceted condition that impacts upon a wide range of functions including speech [4, 5], functional communication [6–8], motor function [9], cognition [10, 11] and quality of life [12, 13]. Levodopa-based pharmacotherapy has become the mainstay of PD treatment in recent decades and has demonstrated good efficacy in targeting motor symptomatology [14, 15]. Nevertheless, there are major limitations associated with this therapeutic regimen, including the risk of developing dyskinesia [16] and the lack of evidence of reliable benefit on speech [17, 18] and non-motor symptoms including cognitive impairment [19].

In light of the limitations of currently available pharmacotherapy, there has recently been increasing interest in and evidence for the potential benefit of non-pharmacological interventions in PD, including the potential benefit of exercise on cognitive impairment [20]. However, the latest Cochrane systematic review on speech-and-language therapy (SLT) for PD [21] was inconclusive regarding the potential benefit of current therapeutic techniques and an updated narrative review [20] found no new studies on SLT in PD since 2013. Therefore, it is worth considering a broader range of interventions that may provide vocal benefit through the conduct of everyday activities. Moreover, since social isolation resulting from difficulties with functional communication is believed to be more problematic for people with PD than physical speech impairment [22], a broader range of interventions that increase social contact may be beneficial both for communication and overall sense ofwellbeing in PD.

There is evidence that participation in the performing arts brings psychosocial and health-related benefits in the general population [23] and in recent years there has been substantial interest in the potential therapeutic benefit of performing arts for people with chronic medical conditions. For example, a meta-analysis of randomised controlled trials (RCTs) published in 2014 [24] found evidence that dance led to clinically meaningful improvements in motor function, balance and quality of life in PD. Music therapy is another potential therapeutic modality and there is evidence from a systematic review for its efficacy in neurological disorders [25], although its benefit in PD may be complicated by impairments of rhythm [26] and melody [27, 28] associated with the condition. However, a recent narrative review [29] provided evidence that music therapy, including the playing of instruments, may provide therapeutic benefit for people with PD.

Singing and in particular the potential benefits of communal singing have in recent years come to the attention of the public, for example in the UK through the popular TV programmes The Choir and Sing While You Work. Community choirs for people with neurological conditions, including PD, have been established in a range of localities. One prominent example is the CeleBRation Choir [30] set up by the Centre for Brain Research at the University of Auckland in New Zealand. Singing as a potential therapeutic intervention in the context of PD does however face an additional barrier in impaired phonation (voicing) [31, 32], which is a fundamental component of singing production [33]. Therefore, it is important to systematically evaluate the evidence base regarding singing as a potential therapeutic intervention in PD.

No systematic review of the potential benefit of singing in PD could be identified. A narrative review in 2010 [34], which covered a spectrum of neurological conditions but included PD, cited two studies [35, 36] that suggested that singing in PD may have a beneficial effect on speech outcomes. Nevertheless, both studies were relatively small uncontrolled studies, which do not provide as reliable evidence for interventional benefit as RCTs, since randomisation addresses the issues of selection bias and confounding that may otherwise arise due to baseline between group differences [37]. Moreover, non-systematic reviews do not state a reproducible methodology and may be subject to selection bias [38]. Therefore, in this article we report the result of a systematic review assessing evidence regarding the benefit of singing on speech and communication, cognitive status, motor function and quality of life in PD.

MATERIALS AND METHODS

We undertook a systematic review, following the PRISMA statement [39], of studies investigating the potential benefit of singing on speech, functional communication, cognitive status, motor function and quality of life in PD. Due to the heterogeneity of intervention characteristics and outcome measures, a narrative systematic review rather than a quantitative meta-analysis was conducted. This study was a systematic review of existing literature and no human or animal participants were recruited. Therefore, no ethical approval was required.

Explicitly stated, our research questions were:

How many studies assessed the potential benefit of singing in PD on each of speech, functional communication, cognitive status, motor function and quality of life?

RESULTS

Study characteristics are presented in Table 1, a synopsis of interventions in Table 2, an overview of study results in Table 3 and the quality assessment results in Table 4. Since there was too much heterogeneity of intervention characteristics and outcome measures to conduct meta-analysis, a narrative review was conducted and results are presented thematically. Below, we provide a concise synthesise of findings on each theme. Further details at the study level are found in Table 3.

DISCUSSION

This is the first systematic review of evidence regarding the potential benefit of singing in PD. It offers an important opportunity to take stock on the current state of research in this field and inform future research directions. The first main finding is that research in this area is limited in terms of the number of studies, sample sizes and study methods. A total of only seven studies were included and all but one was assessed as at high risk of bias. No studies included a control arm or randomised participants to receive or not receive the singing intervention. The second main finding is that research has focused mainly on speech with a much lesser focus on functional communication and wider psychosocial aspects such as quality of life. No included studies addressed potential improvements in cognitive status or motor function, while only one study assessed each of functional communication and quality of life – neither finding evidence of statistically significant benefit. The third main finding is that the majority of studies (five out of seven, 71%) provide some, but not unequivocal, evidence of the benefit of singing on speech in PD across a wide range of speech features.

It is important to consider the methodological limitations of the existing body of literature, so that more robust future studies can be designed that will give a more definitive answer regarding the potential benefits of singing in PD. Of the seven included studies, six (86%) were assessed as at high risk of bias. Only two studies (29%) provided sufficient information on the diagnosis of PD to enable us to have confidence that there is not substantial diagnostic inaccuracy bias in the study sample. The median sample size per study was only 17 (range 4–28) and no study provided suitable justification or rationale for this sample size. Therefore, studies may have been underpowered, which could have biased their findings towards the null [50] and might explain some of the negative and inconclusive findings among studies included in this review. Moreover, potentially related to issues of sample size, we identified concerns over lack of participant representativeness in all seven studies and concerns regarding the sampling strategy of six studies (86%). This sampling bias may limit the generalisability of the findings of included studies. Moreover, in six studies (86%), we identified concerns regarding the suitability of the analysis conducted and/or how it was reported in the included journal articles.

All included studies used a repeated-measures single-group design, although Shih et al [44] attempted to conduct a two-group pilot controlled trial only to abandon this aspect due to recruitment difficulties. The randomised controlled trial (RCT) [37, 52] is the gold standard research design for assessing the efficacy of interventions with numerous benefits including the ability to address issues of selection bias and confounding. Tanneret al [45] justify their choice of a repeated-measures single-group design by citing the extent of between- and within-participant variability associated with PD [53]. However, this evidence is from tasks of a rather specific nature and with interventions among included studies lasting up to two years, the neurodegeneration associated with PD would limit the ability of participants to serve as their own control. The only RCT that we identified [54] was rendered ineligible by the presence of only one person with PD in the singing arm, effectively rendering it a case report for our purposes.

These important limitations notwithstanding, included studies offer some evidence for a benefit of singing in PD on a range of speech features, namely read speech prosody rating, read speech vocal fatigue rating, maximum phonation time, fundamental frequency, fundamental frequency variation, carer-rated intelligibility, intensity, intensity range and three components from the Frenchay Dysarthria Questionnaire [47] – laryngeal pitch, laryngeal volume and laryngeal speech. Therefore, while the evidence is not conclusive, it is sufficient to allow us to say that singing may offer benefit to the speech of people with PD. Moreover, there is more general evidence [24, 54] for the benefit of performing arts including musical approaches as part of a treatment approach in neurological conditions and indeed group-based approaches may reduce social isolation. This is important in context of the finding by Miller et al. that “central concerns were not that voice was monopitch, sound contrasts difficult, or similar ...  Rather individuals focused on the impact changes exerted on ability to communicate [and] their view of themselves” [22]. In essence, wider participation issues including social isolation appear to be of greater importance to people with PD than physical speech difficulties at the International Classification of Functioning Disability and Health (ICF) [55] impairment level.

Therefore, our current provisional suggestion is that clinicians and healthcare providers should consider using singing as part of a rehabilitative programme in PD where they consider it would be useful, drawing on their clinical experience and the experiences of other clinicians who have used such approaches. The organisation of clinical care varies greatly between healthcare systems. In some countries, the multidisciplinary healthcare team may involve a music therapist and/or healthcare service providers may support community performing arts initiatives. On the other hand, in other localities, provision of therapies to improve communication is solely the remit of SLT services. SLT for PD may take many forms. In the UK, many healthcare trusts continue to provide traditional forms of SLT, whereas some trusts fund Lee Silverman Voice Therapy (LSVT) [56], which forms the mainstay of SLT for PD in countries such as the USA. Whereas approaches such as LSVT require a fairly high level of cognitive functioning, singing familiar songs is a less cognitively demanding therapeutic option that is more accessible to those clients with greater cognitive impairment (R.A. Atkinson, personalcommunications).

We have several key recommendations for further research. First, future studies need to use and report more clearly defined diagnostic criteria for PD in order to reduce potential diagnostic bias and allow its evaluation. Second, more effective sampling strategies and greater sample size rationale are required in order to increase generalisability. Third, outcome measures should more often extend beyond ICF impairment level speech alterations and assess other aspects of PD that may be of greater importance to people with the condition and their friends and family members. Fourth, it is important to investigate the optimal frequency, length and intensity of intervention and separate out the effects of singing from those of vocal exercises used as warm-up for singing. Fifth, it is important to assess the generalisability and sustainability of improvements in the longer term beyond the finish of the intervention period. Sixth, RCTs are required in order to address the methodological limitations associated with single-group repeated-measures designs and provide gold standard evidence for the evaluation of the potential benefits of singing in PD.

This review has several notable strengths. First, it provides what is to our knowledge the first systematic review focused on the potential benefit of singing in PD. Second, it offers a systematic approach to inclusion, data extraction, quality assessment and reporting in order to minimise the risk of bias. Third, drawing on the linguistic skills of the authors, articles in seven languages could be considered. Fourth, as per the systematic review by MSB and JB earlier in 2016 [41], we restricted the eligible literature to peer-reviewed journal articles, which we consider reduces the risk of methodological bias, since peer-reviewed journals play a valuable role in validating the quality of research [57]. Moreover, the relatively recent emergence of reputable ‘mega-journals’ addresses issues of publication bias by basing decisions on scientific rigour rather than perceived novelty and impact. It should be noted, however, that some systematic review authors, such as Blackhall and Ker [58], disagree with this stance and favour the inclusion of ‘grey literature’ to ensure complete coverage of the evidence. Fifth, all articles selected for full-text evaluation were successfully retrieved for evaluation.

There are, however, a few limitations to this review that should be taken into account. First, while the range of eligible languages was relatively wide, it was focused on European literature and may not allow complete coverage of the world’s literature. Second, while two reviewers designed searches and conducted all screening, data extraction and quality assessment procedures, only one reviewer conducted the searches. Third, the extent of heterogeneity of intervention characteristics and outcome measures precluded the use of meta-analysis. If the data had been suitable, this would have provided a pooled quantitative estimate of effect size across all included studies.

In conclusion, we present the first systematic review of evidence for the potential benefit of singing for people with PD. The majority of studies provide some evidence for a benefit on speech, although evidence of benefit on other aspects of life in PD is limited. Methodological limitations were commonplace in the identified studies. High quality research including RCTs are required in order to enable us to reach a firm conclusion regarding the scope of the potential benefit offered by singing interventions in the context of PD.

CONFLICTS OF INTEREST

SMB has served as a conductor for Revelation Rock-Gospel Choirs and the SLT Singers, Guy’s and St Thomas’ NHS Foundation Trust. These roles involved promoting the benefits of community singing. MSB has served as a conductor, branch president and national forum committee member for Revelation Rock-Gospel Choirs. MSB received expenses for travel to meetings across the UK as part of this role, which involved promoting performing arts involvement among university students. In the autumn of 1999, SMB undertook a paid internship as a laboratory assistant in the pharmaceutical industry. MSB’s current research is indirectly funded by money from the pharmaceutical companies Pfizer, AbbVie and UCB. These companies manufacture pharmacological therapies for PD, which could be perceived as in competition with non-pharmacological therapies for PD such as those discussed in this article. JB and RAA declare no conflicts of interest.