Abstract
OBJECTIVES: To test associations between subclinical brain infarcts (SBIs) and functional decline independently of intervening clinical vascular events and other vascular risk factors. DESIGN: Longitudinal follow-up for a mean 7.3 years. Generalized estimating equation models were used to test associations between SBIs, number of perivascular spaces (PVSs), baseline Barthel Index (BI), and change in BI, adjusting for sociodemographic, vascular, and cognitive risk factors and for stroke and myocardial infarction occurring during follow-up. SETTING: Population-based prospective cohort study. PARTICIPANTS: Stroke-free individuals from the racially and ethnically diverse Northern Manhattan Study (N=1,290). MEASUREMENTS: Annual functional assessments using the BI (range 0-100). RESULTS: Mean age was 70.6 ± 9.0, 40% of participants were male, 66% were Hispanic, 193 (16%) had SBIs, and 508 (42%) had large PVSs. SBIs were not associated with baseline BI. In a fully adjusted model, there was a change in BI of -0.85 points per year (95% confidence interval (CI)=-1.01 to -0.69); those with SBI had an additional change in BI 0f -0.88 points (95% CI=-1.43 to -0.32). There were no associations between PVS and baseline BI or change in BI. CONCLUSION: In a large population-based study, we found a strong and independent association between "subclinical" markers of cerebrovascular injury and important clinical, person-centered functional trajectories. Future research could clarify the evolution of such subclinical markers over time and test strategies to prevent their progression and minimize related disability. J Am Geriatr Soc 66:2144-2150, 2018.