Background Evidence for longer term exercise delivery for people with Parkinson's disease (PwP) is deficient.

Aim Evaluate safety and adherence to a minimally supported community exercise intervention and estimate effect sizes (ES).

Methods 2-arm parallel phase II randomised controlled trial with blind assessment. PwP able to walk ≥100 m and with no contraindication to exercise were recruited from the Thames valley, UK, and randomised (1:1) to intervention (exercise) or control (handwriting) groups, via a concealed computer-generated list. Groups received a 6-month, twice weekly programme. Exercise was undertaken in community facilities (30 min aerobic and 30 min resistance) and handwriting at home, both were delivered through workbooks with monthly support visits. Primary outcome was a 2 min walk, with motor symptoms (Movement Disorder Society Unified Parkinson's Disease Rating Scale, MDS-UPDRS III), fitness, health and well-being measured.

Results Between December 2011 and August 2013, n=53 (n=54 analysed) were allocated to exercise and n=52 (n=51 analysed) to handwriting. N=37 adhered to the exercise, most attending ≥1 session/week. Aerobic exercise was performed in 99% of attended sessions and resistance in 95%. Attrition and adverse events (AEs) were similar between groups, no serious AEs (n=2 exercise, n=3 handwriting) were related, exercise group-related AEs (n=2) did not discontinue intervention. Largest effects were for motor symptoms (2 min walk ES=0.20 (95% CI −0.44 to 0.45) and MDS-UPDRS III ES=−0.30 (95% CI 0.07 to 0.54)) in favour of exercise over the 12-month follow-up period. Some small effects were observed in fitness and well-being measures (ES>0.1).

Conclusions PwP exercised safely and the possible long-term benefits observed support a substantive evaluation of this community programme.

Trial registration number NCT01439022.