RESUMO
BACKGROUND AND PURPOSE: Levodopa-induced dyskinesia (LID) is a common motor complication in patients with Parkinson's disease (PD). Although amantadine is indicated for LID treatment, it is uncertain whether early treatment with amantadine reduces the risk of LID in patients with PD. We aimed to evaluate the association between amantadine treatment and LID onset in patients with early-stage PD. METHODS: This was a hospital-based retrospective cohort study that used electronic medical records from January 1, 2009 to October 31, 2016. The effect of amantadine on LID onset was compared with those of anticholinergics and monoamine oxidase type B inhibitors in patients with PD. Propensity-score weighting and landmark analysis were used to reduce potential confounding. The time to LID onset was analyzed using Cox models. Sensitivity analyses were performed to determine the robustness of the results. RESULTS: The analyses included 807, 661, and 518 patients at 6-, 12-, and 18-month landmark points, respectively. Amantadine use was associated with delayed LID onset in the 6- and 12-month landmark analyses, with adjusted hazard ratios of 0.65 (95% confidence interval [CI] = 0.49-0.86) and 0.64 (95% CI = 0.47-0.88), respectively. Sensitivity analysis findings were comparable to those of the main analysis. CONCLUSIONS: Early treatment with amantadine may delay LID onset more than treatment with other symptomatic agents. Further studies are needed to elucidate the mechanism of amantadine in LID onset delay and to validate our findings.