Antihypertensive medications and risk of Parkinson's Disease: multi-cohort trial emulation and transcriptomic analyses

Summary Background No known treatments prevent Parkinson's disease (PD) onset or delay its progression. Prior observational studies of antihypertensive medications and PD risk have produced inconsistent findings and may be susceptible to confounding and reverse causation. Methods We applied a multi-stage analytical framework integrating medication-wide screening, independent cohort replication, target trial emulation, and exploratory transcriptomic analyses. Medication-wide association screening was conducted in a large-scale electronic health record (EHR)-linked biobank using logistic regression, with replication in independent cohorts. For candidate drug classes, we emulated a target trial among individuals with primary hypertension initiating first-line antihypertensive therapy in the Merative MarketScan Medicare Database. Incident PD was the primary outcome, and associations were estimated using inverse-probability-of-treatment-weighted Cox proportional hazards models. Exploratory transcriptomic analyses integrated single-cell drug-perturbation data with PD brain expression datasets. Findings Across medication-wide screening and independent cohort replication, losartan (an ARB) and amlodipine (a CCB) were consistently associated with lower odds of PD. The target trial emulation included 269,617 individuals with primary hypertension (25,965 ARB initiators; 36,873 CCB initiators; and 206,779 initiators of other first-line antihypertensives). ARB initiation was associated with a lower hazard of incident PD than initiation of other first-line antihypertensives (adjusted HR [aHR] 0.70, 95% CI 0.59-0.83); the corresponding association for CCB initiation was smaller (aHR 0.88, 95% CI 0.77-0.99). The ARB association was generally consistent across sensitivity analyses addressing alternative exposure definitions, measured prodromal PD characteristics, and comparator choice, but was attenuated at a 5-year PD outcome lag. ARB initiation was also associated with a lower PD risk than ACE inhibitor initiation (aHR 0.81, 95% CI 0.67-0.97). Exploratory transcriptomic analyses showed directional opposition between PD-associated molecular signatures and transcriptional changes after sacubitril/valsartan exposure. Interpretation ARBs initiation was associated with a lower risk of incident PD than initiation of other first-line antihypertensive medications. The attenuation observed for 5-year lag time leaves room for alternative explanations including the influence of prodromal disease on ARB use, delayed ascertainment, and other possible sources of bias. These findings warrant prospective evaluation but do not endorse ARB use for the clinical prevention of PD.