Group-based music intervention in Parkinson's disease - findings from a mixed-methods study.

The study was underpowered, and missing values were 4%-17%, so null findings may be less clear. Only 46 patients completed the study, and analysis was per-protocol rather than intention-to-treat. Five dropouts were all women, which may limit generalizability. Medication changes were not included in the statistical model, and patients were not re-tested at exactly the same time of day. Many participants had cognitive impairment at baseline (23 patients, 50%), which may have made complex dual-task learning harder. The primary dual-task test may not have matched the therapy, and the control group received no competing activity, so expectancy effects cannot be separated. The intervention combined music, movement, cognitive tasks and social group activity, so the contribution of any single component cannot be separated. Qualitative reports of feeling better or having fun are experiences, not evidence that the intervention improves clinical outcomes.

No funding source is stated in the supplied text. One author was a non-practicing certified practitioner of the Ronnie Gardiner Method, but was blind to outcome results and did not take part in interviews.

Do not conclude that music alone improves motor-cognitive dual-tasking, balance, cognition or freezing of gait. The primary dual-task walking outcome was not significantly different from usual care, and the intervention combined music, movement, cognitive tasks and social group activity, so the effect of any single component cannot be separated.