Dancing with Parkinson's disease: feasibility randomised controlled trial

Background: Self-help and physical leisure activities are important to maintain safe functional mobility among older people. Purpose: The purpose of this study was to determine the feasibility of conducting a phase III randomised controlled trial (RCT) to evaluate the benefits of dance among people with Parkinsons’ (PwP). We built on previous proof-of-principle trials, focusing on areas of greatest methodological uncertainty. Methods: Ethical approval was granted by the National Research Ethics Service (NRES) Committee South Central Southampton A, reference: 12/SC/0355. We screened 103 PwP and fifty-one were recruited and randomised to the control (n = 15) or experimental group (n = 36). Three ballroom (Social Foxtrot, Waltz, Tango) and 3 Latin American (Rumba, Cha Cha, Rock ‘n’ Roll) dances were taught by professional teachers in a dance centre. The classes lasted one hour, twice a week, for 10 weeks and PwP danced with their spouse, a friend or a volunteer. A battery of assessments (balance, confidence, spinal posture, mobility and health outcome) was completed at home by an assessor (unaware of participant group allocation) at baseline, 3 and 6 months post-randomisation. Results: The average age of participants was 71 years (49-85); 25 were male. Partners and volunteers were of similar age. The feasibility findings focused on recruitment, retention, outcome measures and dance selection. Target recruitment was achieved, with most PwP coming from Parkinsons’ UK Groups. Of the 35 people who took part in the dance classes, 19 nominated a dance partner and 16 danced with healthy volunteer dance partners (recruited by the study team). Five people did not complete the dance classes for health and personal reasons, with one disliking dancing. The remaining participants completed on average 18 of the 20 classes. Travel assistance was costly and time-consuming but important to participants. One data set was unavailable at 3 months and 4 data points were missing from the assessment data. Outcome variables were compared between dance and control groups at 3 and 6 months separately, in analyses of covariance. Estimated mean differences between dance and control (controlled for baseline differences) were explored with model based 95% confidence intervals. As expected, no overall significant differences were found in this feasibility study. Subgroup analysis identified that (a) previous fallers, (b) PwP who have had the condition for longer (diagnosed = 2011) and (c) male participants appeared to benefit most from the dance intervention. Findings suggested trends and significant improvements (p Conclusion(s): We have demonstrated the feasibility of conducting a RCT to evaluate the benefits of ballroom and Latin American dancing for PwP, with procedural recommendations for the main trial. Implications: More research is needed.