P4‐031: Systolic function is associated with brain aging: The Framingham Heart Study

Recent work has related left ventricular (LV) systolic dysfunction to increased white matter changes and executive dysfunction in aging adults with prevalent cardiovascular disease (CVD). These associations are theoretically due to chronic reductions in systemic blood flow disrupting cerebral blood flow homeostasis and contributing to subclinical brain injury. To assess the generalizability of prior findings, we examined measures of LV systolic function (cardiac output and ejection fraction) to neuropsychological and neuroimaging variables associated with brain aging among a large cohort of community dwelling adults. We hypothesized that LV systolic function would be associated with brain magnetic resonance imaging (MRI) and neuropsychological markers of pre-clinical cerebrovascular changes (e.g., executive functioning, white matter hyperintensities (WMH)) and Alzheimer's disease (e.g., memory, brain volume). Brain MRI, cardiac MRI, and neuropsychological data were collected on 1505 Framingham Heart Study Offspring Cohort participants free from clinical stroke, TIA, or dementia (34–84 years, 61±9; 54% women). Brain aging variables included five neuropsychological factors (verbal memory, visuospatial memory, verbal learning, information processing/executive functioning, and language/perception) and two brain MRI variables (WMH and total brain volume). Brain aging variables were related to cardiac MRI-assessed LV ejection fraction and cardiac output, adjusted for age, sex, systolic blood pressure (BP), smoking status, fasting glucose, diabetes, BP treatment, atrial fibrillation, prevalent CVD, and time to brain MRI. Models relating cardiac output to brain aging variables included body surface area as a covariate. In multivariable models, cardiac output was positively associated with total brain volume (standardized β=0.07, p=0.02). LV ejection fraction was inversely related to WMH (standardized β=-0.05, p=0.04). Neither systolic function variable was associated with any of the neuropsychological factors (all p-values>0.12). Our analyses demonstrate that measures of systolic function are associated with neuroimaging variables of brain aging, including white matter changes and brain volume. Whether these relations are causal is unknown given our cross-sectional methodology. However, we propose that subclinical systolic dysfunction may be associated with maladaptive brain aging, which is supported by the current findings and past research using clinical-referral samples. Future longitudinal work will clarify whether these associations are causal.