Midlife Vascular Risk Factors, Parkinson Disease, and Parkinson Disease-Dementia: The ARIC Study

Vascular risk factors (VRFs) such as smoking, hypertension, obesity, and diabetes are associated with dementia, but their importance in Parkinson disease (PD) and PD-dementia (PDD) is less well understood. Previous studies demonstrated that smoking may be protective of PD, but its role in PDD is unclear. The primary objective was to examine the association between midlife VRFs and the risk of developing PD and PDD in older adults.

The prospective, longitudinal community-based cohort Atherosclerosis Risk in Communities study recruited Black and White adults (45-64 years old) from Jackson, MS; Forsyth County, NC; Minneapolis suburbs, MN; and Washington County, MD, in 1987-1989. VRFs (smoking status, hypertension, obesity, hypercholesterolemia, and diabetes) were measured at the baseline visit. PD cases were retrospectively adjudicated through 2016 by reviewing participant medications, self-reported physician-made diagnoses, and hospitalization and death surveillance data. Dementia cases were adjudicated using in-person and telephone-based cognitive testing, informant interviews, and hospitalization codes. PDD cases were defined as an adjudicated PD diagnosis followed by an adjudicated dementia diagnosis. Cox proportional hazard models, adjusted for age, race, sex, education level, and APOE ε4 status, evaluated midlife VRFs together in association with PD, with separate models evaluating PD without dementia, PDD, and dementia without PD.

Among 13,875 individuals with nonmissing VRF and outcome data with up to 30 years of follow-up (25% Black, 54% female), 179 developed PD at a mean age of 73.4 years, 94 developed PDD at a mean age of 79.2 years, and 1,791 developed dementia without PD at a mean age of 79.7 years. Midlife current smoking (hazard ratio [HR] 0.36, 95% CI 0.21-0.61) was significantly associated with a lower PD rate, independent of other risk factors and demographics. Midlife current smoking was also significantly associated with a lower rate of PDD (HR 0.41, 95% CI 0.18-0.95). Other VRFs were not associated with either PD or PDD.

Smoking status in midlife was associated with lower rates of PD and PDD, but other VRFs had no association with PD or PDD. Further studies should evaluate changes of these VRFs over the life course and explore mechanisms for the observed associations.