Objectives: The effect of rhythmic auditory stimulation (RAS) on gait velocity, cadence, stride length, and symmetry was studied in 31 patients with idiopathic Parkinson's disease, 21 of them on (ON) and 10 off medication (OFF), and 10 healthy elderly subjects.

Method: Patients walked under four conditions: (1) their own maximal speed without external rhythm; (2) with the RAS beat frequency matching the baseline cadence; (3) with RAS 10% faster than the baseline cadence; (4) without rhythm to check for carry over from RAS. Gait data were recorded via a computerised foot switch system. The RAS was delivered via a 50 ms square wave tone embedded in instrumental music (Renaissance style) in 2/4 metre prerecorded digitally on a sequencer for variable tempo reproduction. Patients on medication were tested in the morning 60-90 minutes after medication. Patients off medication were tested at the same time of day 24 hours after the last dose. Healthy elderly subjects were tested during the same time of day.

Results: Faster RAS produced significant improvement (P < 0.05) in mean gait velocity, cadence, and stride length in all groups. Close synchronisation between rhythm and step frequency in the controls and both Parkinson's disease groups suggest evidence for rhythmic entrainment mechanisms even in the presence of basal ganglia dysfunction.

Conclusions: The results are consistent with and extend prior reports of rhythmic auditory facilitation in Parkinson's disease gait when there is mild to moderate impairment, and suggest a technique for gait rehabilitation in Parkinson's disease.