Best Exercise for Parkinson's Disease: What New Review Says
11 Oct 2026
Almost every type of exercise helped people with Parkinson's compared with doing nothing, and no single exercise has been proven best. A new review of 173 trials found mind-body exercise gave the largest average gain in movement symptoms, and that more weekly exercise was linked with better results, but the certainty of the evidence was low [1]. The most useful plan is one you can do regularly and safely.
Best Exercise for Parkinson's Disease at a glance
- New review: 173 randomised trials of exercise in Parkinson's, published in October 2026 [1].
- Mind-body exercise (such as tai chi or yoga-style practice) showed the largest average effect on movement symptoms [1].
- Cochrane's 2024 review found dance and gait, balance and functional training probably help movement symptoms [2].
- More weekly exercise was linked to better movement, balance and Timed Up and Go results, but not walking speed [1].
- Confidence in all of this was low or very low, so no exercise or dose can be called ideal [1].
What did the new review actually find?
Researchers pooled 173 randomised trials and compared exercise types with people who did no exercise. They measured movement symptoms (the UPDRS III score), balance, walking speed and the Timed Up and Go test, which times how long you take to stand, walk three metres, turn and sit [1].
Mind-body exercise had the largest average effect on movement symptoms. They also looked at how much exercise people were planned to do each week. Within the same type of exercise, trials with more planned weekly exercise tended to report larger gains in movement symptoms, balance and the Timed Up and Go test, but not in walking speed [1].
Does that mean mind-body exercise is the best?
Not quite. The authors were careful here. When they calculated the range where a new trial's result would be expected to fall, every exercise type crossed zero, and their confidence ratings were low or very low. They state that the evidence did not establish a universally best type, an optimal dose, or a cause-and-effect target for any individual [1].
So the result is a hint, not a prescription. A tai chi class that you enjoy and attend twice a week is likely to do more for you than a "best" programme you drop after three weeks.
What did earlier reviews say?
The Cochrane review of 154 trials (7,837 people, mostly with mild to moderate disease) is the most cited summary. It found moderate-certainty evidence that dance and gait, balance and functional training probably have a moderate benefit on movement symptoms, and that multi-domain training probably has a small benefit. Aqua-based training probably improved quality of life [2].
A separate network analysis of 199 trials ranked exercises for balance. Treadmill training with body-weight support, aquatic exercise, Pilates and cued balance training came out on top for different balance measures, though the confidence was often low [3].
Notice that the three reviews do not agree on one winner. They agree on something simpler: moving regularly beats staying still.
| Review | Exercise that stood out | Certainty |
|---|---|---|
| Wang 2026, 173 trials [1] | Mind-body exercise for movement symptoms | Low or very low |
| Cochrane 2024, 154 trials [2] | Dance; gait, balance and functional training | Moderate for movement symptoms |
| Balance analysis, 199 trials [3] | Body-weight-supported treadmill, aquatic exercise, Pilates | Often low |
Exercise types that stood out in each review [1,2,3].
How much exercise is enough?
The new review linked an extra 250 MET-minutes of structured exercise per week with a modest extra benefit. A MET-minute is a way of counting both time and effort. The estimates were imprecise, and the walking-test result changed when the highest-dose trials were removed [1].
The review does not prove any particular weekly amount is ideal. If you are new to exercise or have had falls, start low and build up with your physiotherapist's guidance.
Is exercise safe with Parkinson's?
In the Cochrane review, only 85 of the trials reported safety data. No adverse events occurred in 40 of them. Where events were reported, falls were the most common [2]. This is why supervision matters if your balance is affected.
- Tell your neurologist before you start something new.
- Choose a class or a physiotherapist who knows Parkinson's, especially if you freeze or fall.
- Ask your doctor when in the day to exercise if your medicine wears off.
- Stop and seek advice if you get chest pain, fainting or new weakness.
Getting help in Delhi
At Brain Bridge Clinic in Safdarjung Enclave, New Delhi, Parkinson's care begins with a structured neurological assessment of tremor, stiffness, slowness, balance and sleep. Based on that, the team can draw up a plan with you and your family, which can include supervised physiotherapy and exercise matched to your balance and stamina. We work alongside your treating neurologist, and medicine decisions stay with them. Bring your current prescriptions and any scan reports. If you would like an assessment, you can book through the website or by phone.
Related pages: Parkinsons, Physical Therapy, Parkinsons.
Frequently asked questions
Which exercise is best for Parkinson's disease?
No single type has been proven best. Dance, gait and balance training, and mind-body exercise all show benefit in reviews, with varying certainty [1,2].
Can exercise slow Parkinson's progression?
The reviews here measured symptoms such as movement and balance after exercise, not long-term progression, so they cannot answer this [1,2].
How often should I exercise?
The new review linked more planned weekly exercise with better results, but it could not name an ideal amount [1]. Ask your physiotherapist to set a level that suits your fitness.
Is walking enough?
Walking is a good base, but balance and strength work matter too. The Cochrane review found gait, balance and functional training probably help movement symptoms [2].
Can I exercise if I have had falls?
Yes, but start under supervision. Falls were the most commonly reported adverse event in trials [2].
References
- Wang L, Wang H, Hong F, et al. Exercise modality and planned weekly exercise exposure for motor outcomes in Parkinson's disease: a systematic review, Bayesian network meta-analysis, and meta-regression. Journal of neurology. 2026;273(11). doi:10.1007/s00415-026-14189-y. PMID 42859236.
- Ernst M, Folkerts AK, Gollan R, et al. Physical exercise for people with Parkinson's disease: a systematic review and network meta-analysis. The Cochrane database of systematic reviews. 2024;4(4):CD013856. doi:10.1002/14651858.CD013856.pub3. PMID 38588457.
- Qian Y, Fu X, Zhang H, et al. Comparative efficacy of 24 exercise types on postural instability in adults with Parkinson's disease: a systematic review and network meta-analysis. BMC geriatrics. 2023;23(1):522. doi:10.1186/s12877-023-04239-9. PMID 37641007.