Singing interventions for people living with Parkinson's: a systematic review and meta-analysis

To evaluate the effects of singing interventions on well-being, mental health and communication, motor and respiratory functions through meta-analysis and to examine the practices used in the singing interventions.

Systematic review and meta-analysis.

Both randomised and non-randomised studies, involving participants living with Parkinson's and receiving singing interventions.

Four databases (CINAHL, Medline, PsycINFO, Web of Science) and Google Scholar were searched. The last search was conducted on 3 April 2025.

Eligible studies reported on the following outcomes: quality of life, voice-related acoustic measures, respiratory function, mental health and motor function. The risk of bias was assessed using the Downs and Black Quality checklist for controlled studies and national heart, lung, blood institute study quality assessment tool for non-controlled studies.

Meta-analyses were conducted to pool effect sizes across included studies using random-effects models. All analyses were performed using Meta-Essentials. Additionally, key elements of singing practices were narratively synthesised.

23 studies (3 randomised controlled trials (RCTs), 20 non-RCTs) involving 540 participants were included in the analysis. Common elements of singing intervention practices included breathing exercises, vocal warm-ups and singing participants' preferred songs, which are largely led by music therapists.Three meta-analyses based on RCTs were conducted on clinical outcome measures; results suggest that singing was favoured in only one outcome measure, vocal loudness during sustained vowel production (standard mean difference (SMD)=0.67, 95% CI 0.29 to 1.05, I²=0%). However, the certainty of this evidence is very low due to a high risk of bias, imprecision and indirectness. Further, when combining results from RCTs and non-RCTs, positive changes for three further outcome measures were observed: maximum phonation time (k=11, n=157, SMD=0.38, 95% CI 0.18 to 0.59, I²=23.48%), vocal loudness of sustained vowel (k=8, n=99, SMD=0.50, 95% CI 0.14 to 0.86, I²=50.96%) and a respiratory function measure of maximal inspiration pressure (k=4, n=65, SMD=0.46, 95% CI 0.07 to 0.85, I²=0%). However, these findings are largely based on non-controlled studies with very low quality of evidence.

Singing interventions may support people living with Parkinson's, but due to insufficient high-quality evidence, we are unable to determine the effects of singing interventions. We discuss implications for future work and practice, emphasising that more robust RCTs are needed.