Improved cognitive flexibility after high-intensity interval training (HIIT) in individuals with Parkinson's disease: a pilot randomized controlled trial.

Twenty people with Parkinson's disease (Hoehn and Yahr stages I–IV, no severe dementia) were recruited at Lucerne Cantonal Hospital in Switzerland and assigned to one of two groups using a randomization list generated by the Blockrand tool and concealed from the treating investigator. The HIIT group cycled on a stationary bike three times a week for three weeks (nine sessions total): five minutes warm-up at 60% of maximum heart rate, then twenty minutes of alternating 30-second bouts at 75% and 90% of maximum heart rate (20 cycles), then five minutes cool-down. The continuous-training group cycled at a steady 85% of maximum heart rate for the same 20-minute workout, with the same warm-up and cool-down. Both groups trained at very similar overall intensities. All sessions were supervised by a physical therapist. Cognitive tests (Trail Making A and B, Stroop I, III, IV) and secondary measures (Mini-BESTest, PDQ-39, FOG-Q, UPDRS II and III, cardiopulmonary exercise testing) were done at baseline and immediately after the final session by the same investigator, who was not blinded. The sample size of 10 per group was chosen to estimate effect sizes for a future larger trial.

The main question — whether HIIT produces greater cognitive improvement than continuous training — was answered with one positive and four null results across the five cognitive tests. Trail Making Test Part B (cognitive flexibility) was the only test showing a significant between-group advantage for HIIT: the HIIT group improved significantly while the continuous group did not, and the interaction reached significance. Stroop III (inhibitory control) showed a significant within-group improvement in the HIIT group but no significant between-group difference. Trail Making A, Stroop I, and Stroop IV showed no significant changes in either group. On all secondary outcomes — balance, motor function, quality of life, freezing of gait, and aerobic capacity — both groups improved or stayed the same with no between-group differences. Notably, VO2max improved significantly only in the continuous-training group, not in the HIIT group, likely because three weeks was too short for the HIIT group to show an aerobic advantage.

This is a PT-delivered cycling programme, and the one positive signal — a small advantage for interval over steady-state cycling on a cognitive flexibility test — is too preliminary to change how you prescribe exercise on Monday. What it does support is the idea that a 3-week, three-times-weekly stationary-bike programme is feasible and well tolerated in PwPD up to Hoehn and Yahr stage IV, and that both interval and continuous formats produce comparable motor and balance gains over that short window. If you are already cycling PwPD patients, this does not tell you to switch to intervals; it tells you a larger trial is warranted before you do.

The cognitive outcomes this study targets — executive function, cognitive flexibility, inhibitory control — sit squarely in the OT domain for PwPD, and the authors note these are strongly associated with functional independence. The single positive signal on Trail Making B is too small and too preliminary to act on, but it is a reason to watch for a larger trial before dismissing aerobic exercise as a component of a broader OT programme aimed at executive function in PwPD.

Ten participants per group is very small; the authors themselves call this exploratory and note the limited statistical power increases the risk of missing a real effect. The assessor was not blinded to group allocation, so cognitive scoring could have been influenced by knowing which group the patient was in. Three weeks is a short intervention; the authors acknowledge this may have been insufficient to produce larger cognitive or aerobic adaptations. Several participants could not complete some cognitive tests (TMT-B, Stroop III, Stroop IV), leaving unequal group sizes of 7–10 for those outcomes and introducing missing-data bias. Baseline cognitive ability differed between groups, as shown by a wider range of MoCA scores in the HIIT group, which makes between-group comparisons less clean. The study was conducted at a single site in Switzerland, limiting generalisability.

Funded by the Jacques and Gloria Gossweiler Foundation. The authors state the sponsor had no involvement in the research that could have influenced the outcome.

Do not read the Trail Making B result as proof that interval cycling is superior to steady-state cycling for cognition in Parkinson's disease. It was the only one of five cognitive tests to show a between-group difference, the group sizes for that test were 8 and 9 (not 10), the assessor was not blinded, and the Stroop III improvement in the HIIT group did not reach between-group significance. A single positive interaction in a 20-person, 3-week pilot is a hypothesis-generating signal, not a treatment recommendation.