Factors predictive of the development of Levodopa-induced dyskinesia and wearing-off in Parkinson's disease

The Stalevo Reduction in Dyskinesia Evaluation in Parkinson's Disease (STRIDE-PD) study compared the initiation of levodopa (l-dopa) therapy with l-dopa/carbidopa (LC) versus l-dopa/carbidopa/entacapone (LCE) in patients with Parkinson's disease. In the current study, the STRIDE-PD study population was investigated to determine the effect of l-dopa dose and other risk factors on the development of dyskinesia and wearing-off. Patients were randomized to receive LCE (n=373) or LC (n=372). Blinded assessments for dyskinesia and wearing-off were performed at 3-month intervals for the 134- to 208-week duration of the study. The patients were divided into 4 dose groups based on nominal l-dopa dose at the time of onset of dyskinesia (or at study conclusion if there was no dyskinesia): group 1, 600 mg/day (n=77). Similar analyses were performed with respect to wearing-off and any motor complication. The times to onset and frequency of dyskinesia, wearing-off, or any motor complication were compared using the log-rank test (overall trend test) and a Cox proportional hazards model (pairwise comparisons). A stepwise Cox proportional hazards model was used to screen predictive factors in a multivariate analysis. The risk of developing dyskinesia and wearing-off increased in an l-dopa dose-dependent manner (P