Physiotherapy interventions for balance impairments in Parkinson’s disease: evidence from a systematic review and dose-response meta-analysis

Balance impairments are among the most disabling symptoms of Parkinson and are often poorly responsive to pharmacological therapy. Physiotherapy (PT) is a key non-pharmacological intervention to improve postural control, though optimal exercise characteristics and dose remain unclear. We conducted a systematic review and dose-response meta-analysis to evaluate PT's effectiveness on balance in people with Parkinson's disease (PwPD) and define a dose-response relationship. A systematic search of PubMed/MEDLINE, EMBASE, and Web of Science was done from March-April 2024 and updated in February 2025. Eligibility criteria comprised RCTs or cross-over studies included adults with idiopathic PD receiving exercise targeting balance. Two reviewers independently assessed risk of bias using the Cochrane Risk of Bias tool for randomized trials (RoB 2.0). Thirty studies (n participants = 2,932; mean ± SD age = 69.3 ± 4.0 years; mean ± SD disease duration = 6.7 ± 1.6 years; mean ± SD Hoehn & Yahr = 2.3 ± 0.5) were included. Balance outcomes were Mini-Balance Evaluation Systems Test (Mini-BESTest), Timed Up and Go (TUG), Berg Balance Scale (BBS), Unified PD Rating Scale (UPDRS-III), and Sensory Organization Test (SOT). Physiotherapy led to moderate balance improvements (SMD = 0.56, 95%CI 0.38-0.74), with balance-specific training showing the greatest effect (SMD = 0.64, 95%CI 0.35-0.93). No clear linear dose-response was found between training volume and effect size. Subgroup analyses confirmed consistent benefits across Mini-BESTest, BBS, and TUG. These findings support personalized, balance-focused rehabilitation in PD and highlight the need for clearer intervention protocols.