OBJECTIVE:

To evaluate the influence of long-term zolpidem use on the incidence of developing Parkinson'sdisease.

METHOD:

2,961 subjects who used zolpidem for the first time longer than 3 months between 1998 and 2000 were identified in the National Health Insurance system of Taiwan. Subjects without a history of zolpidem use were randomly selected as a comparison cohort and frequency matched to zolpidem users based on age, sex, and index date. The diagnosis of Parkinson's disease was based on the criteria of the International Classification of Diseases, Ninth Revision, Clinical Modification. Its incidence until the end of 2009 was calculated and its hazard ratios (HRs) and 95% CIs were estimated using Cox proportional hazards regression models and Kaplan-Meier analysis.

RESULTS:

The overall incidence of Parkinson's disease was greater among zolpidem users than in the comparison cohort (HR = 1.88; 95% CI, 1.45-2.45). However, there was no difference in Parkinson's diseaseincidence between these 2 cohorts after 5 years of observation. The risk of Parkinson's disease increased with increasing zolpidem dose, with an HR of 0.70 for low-dose users (< 400 mg/y) and 2.94 for high-doseusers (≥ 1,600 mg/y). The incidence of Parkinson's disease was greater in subjects using zolpidem only (HR = 2.35; 95% CI, 1.66-3.33) compared to those using benzodiazepines only (HR = 1.31; 95% CI, 0.91-1.90). By stratified analysis, zolpidem use with concurrent depression (HR = 4.79) increased the risk of Parkinson'sdisease compared to that of zolpidem users without concurrent depression.

CONCLUSIONS:

Zolpidem use might unmask preclinical Parkinson's disease, especially in patients with depression. However, large population-based, unbiased, randomized trials are warranted to confirm this finding.