Objective: To evaluate the association between time-varying statin use and PD risk in a large-scale population-based study.

Background: Statins have been shown to downregulate neuroinflammation, associated with development of Lewy-bodies in Parkinson’s disease (PD). However, previous epidemiological studies,considering a time-fixed statin use, have yielded conflicting results.

Methods: Using BIG-DATA of Maccabi-Healthcare-Services (MHS-covers 25% of the Israeli population) we established a cohort of stain-users (ages 40-79) firstly purchased statins during 1.1.99-31.12.12 and underwent low density lipoprotein cholesterol (LDL-C) blood tests. Time-varying statin use was measured by proportion of days’ cover (PDC, 0-100%) for each year of the follow up period (FUP), from 1st purchase until end point (date of PD diagnosis, death, leave MHS or end of study, whichever occurred first). PD assessment was based on a validated anti-Parkinsonian drugs tracer approach. We used time-dependent Cox models (time scale=FU) to associate between time-varying statin use and PD risk 1-year lagged, adjusted to age at 1 st purchase and changes in LDL-C levels. Analysis was stratified by sex, LDL-C levels at baseline (≤130, >130 mg/dl) and age categories to account for possible interactions.

Results: The cohort of statin users included 232,877 individuals (49.3% men), who had at least 3 purchases during study period. Mean age at 1st purchase was 56.5 (±9.8) years among men and 58.7 (±9.2) years among women. PDC distribution for the whole FUP differed between men and women: medians 58.3% and 54.1% respectively, with the same IQR (25%-83%). During a mean FU of 7.7 (±3.4)years, 2,550 (1.1%) incident PD cases were identified. In the 1-year lagged analysis, we did not find an association between the dynamics in statin use (all statins) and PD risk (all HRs close to unity). Similar results were found in a refined analysis regarding statin type and statin efficacy. A sensitivity analysis on individuals who used simvastatin only (72% of all purchases) yielded the same null results.

Conclusions: Our big-data analysis suggests that time-varying statin use is not associated with PD risk in contrast to some previous reports that measured a time-fixed statin use throughout the entire FUP. Further studies in this subject should use large-scale cohorts and refined assessment of dynamics in statin use.

« Back to 2016 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/dynamics-in-statin-use-low-density-lipoprotein-cholesterol-levels-and-parkinsons-disease-a-population-based-cohort-study-using-big-data/