Results
The mean age of persons with PD and their controls were 73.5 years and 73.6 years, respectively (range 44.4–95.8), and majority of them were men (table 1). Median duration of diabetes on index date was 8.8 years in persons with PD and 8.5 years in persons without PD. There were no differences in comorbidities, except for history of traumatic brain injury which was more prevalent among PD cases compared to controls.
Statin users were more likely to be men (61.9% of users and 38.1% of nonusers) (table 2). All comorbidities, except cancer and rheumatoid arthritis, were more common among statin users, with largest difference observed in cardiovascular diseases (71.9% of users and 59.7% of nonusers).
Prevalence of any statin exposure before or only during the three-year lag was similar among persons with PD and without PD (54.2% vs. 54.4% and 10.9% vs. 10.6%) (table 3). Simvastatin was the most commonly used statin, with 38.8% of cases and 40.1% controls exposed before the three-year lag. There was no difference in use of specific statins between PD cases and controls (table 3). Use of hydrophilic statins was not common before the three-year lag (n=901), and majority of their users had also purchased lipophilic statins (78.0%, n=703). There was no difference in prevalence of lipophilic or hydrophilic statin use between cases and controls.
Any statin use before the three-year lag was not associated with the risk of PD (adjusted odds ratio (aOR) = 1.02; 95% CI: 0.91–1.15) (table 4). In the dose-response analyses, an increased risk of PD was observed in the highest statin exposure tertile compared to statin non-use (aOR = 1.21; 95% CI: 1.04–1.42). Similarly, in the secondary analysis restricted to statin users only, an increased risk of PD was observed in the highest exposure tertile compared to the lowest exposure tertile (aOR = 1.29; 95% CI: 1.07–1.57).