Abstract
Despite existing evidence linking dyskinesia to levodopa, the primary treatment for Parkinson's, the dose-response relationship and risk factors remain uncertain. In this study, the risk for dyskinesia in patients with Parkinson's disease receiving levodopa was evaluated via meta-analysis and meta-regression approaches to examine dyskinesia risk factors more reliably and improve treatment strategies and patient care. The PubMed and Embase databases were searched to identify randomized controlled trials comparing levodopa with other anti-Parkinson's drugs published in English before June 31, 2023. The primary outcome was dyskinesia, and a risk of bias assessment was performed. In total, 24 studies met the inclusion criteria; 21 had a low risk of bias, and 3 had a high risk of bias. These studies included 4698 patients with Hoehn and Yahr Grade I-III Parkinson's disease. Our meta-analysis showed that the risk of dyskinesia was higher for levodopa than for other anti-Parkinson's drugs (odds ratio: 2.52 [95% confidence interval: 1.84-3.46]). Dyskinesia was not related to age (slope coefficient: 0.185 [0.095]; P = 0.061), disease duration (slope coefficient: 0.011 [0.018]; P = 0.566), or treatment duration (slope coefficient: 0.008 [0.007]; P = 0.216). The mean levodopa equivalent dose (slope coefficient: 0.004 [0.001]; P = 0.001) in the experimental group and the differences in drug doses between the experimental and control groups were correlated with the risk of dyskinesia. Results of randomized controlled trials supported an association between the levodopa dose and dyskinesia in patients with Parkinson's disease. Compared with levodopa users, users of other anti-Parkinson's drugs had a lower incidence of dyskinesia. Age, disease duration, and treatment duration were not correlated with dyskinesia. These findings suggest that anti-Parkinson's drugs other than levodopa, particularly in cases of early-stage Parkinson's disease, should be considered to reduce the risk of dyskinesia.