PURPOSE

This retrospective cohort study used a population-based dataset to test the risk for parkinsonism in patients receiving flunarizine and cinnarizine, compared with matched controls.

METHODS

Data were obtained from the National Health Insurance Research Dataset of Taiwan. Patients receiving flunarizine or cinnarizine for more than 1 month between 2000 and 2005 were enrolled. Exclusion criteria included receiving flunarizine, cinnarizine, or antipsychotics for more than 1 month during 1997-1999, a history of neurodegenerative diseases, and an age of less than 30 years. One matched control for each patient was selected. Each participant was followed for diagnosis of parkinsonism within a 3-year observation period. Stroke, diabetes mellitus, total prescription days, and doses of flunarizine or cinnarizine were recorded.

RESULTS

The study and control groups consisted of 9830 subjects. In the study group, 280 patients (2.9 %) were diagnosed with parkinsonism with a median observation period of 1.2 years, and 49 participants (0.5 %) were diagnosed in the control group with a median observation period of 1.9 years. The adjusted hazard ratio for parkinsonism among patients receiving flunarizine and cinnarizine was 5.117 (95 % CI = 3.758-6.967). Age, stroke, and diabetes mellitus were significant risk factors, but female sex and total doses of the study drugs were not.

CONCLUSIONS

This study demonstrates that flunarizine and cinnarizine significantly increase the risk for parkinsonism. The treatment benefits of these two agents should be balanced with this adverse effect. Physicians must look carefully for early signs of parkinsonism in patients treated with flunarizine and cinnarizine.