Supplementary Material for: Prevalence of cardiovascular drugs and oral anticoagulants use among persons with and without Parkinson’s disease

Introduction
Cardio- and cerebrovascular diseases are common among persons with Parkinson’s disease (PD), but it is unknown how prevalence of cardiovascular drugs and oral anticoagulants use changes in relation to PD diagnosis.

Methods
We investigated prevalence of cardiovascular drugs and oral anticoagulants use among persons with and without PD among 17 541 persons who received incident PD diagnosis in 2001-2015 in Finland and their 116 829 matched comparison persons. Prevalence was calculated in six-month time windows from five years before to five years after PD diagnosis (index date) and compared to a matched cohort without PD using generalized estimating equations.

Results
Persons with PD had higher prevalence of any cardiovascular drugs (unadjusted OR = 1.15; 95% CI 1.11–1.18), and oral anticoagulants (unadjusted OR = 1.16; 95% CI 1.11–1.22) before index date than those without PD. After index date, persons with PD had lower prevalence of cardiovascular drugs (0.94; 95% CI 0.91–0.96), and no difference was observed for oral anticoagulants. Prevalence of any cardiovascular drugs on index date were 66% and 61% for persons with and without PD respectively. Beta blockers were the most common cardiovascular drugs in both cohorts. Warfarin was the most common oral anticoagulant, but use of direct oral anticoagulants increased during the last years of follow-up.

Conclusion
Orthostatic hypotension and weight loss likely explain the decreased cardiovascular drugs use after PD diagnosis. Results with oral anticoagulants may reflect clinical assessment of benefits being larger than risks despite of risks associated with their use in persons with PD.