Subclinical vascular risk composites and dementia imaging biomarkers 17-20 years later in the Multi-Ethnic Study of Atherosclerosis (MESA).
Marc D Rudolph, Jordan Tanley, Jingzhong Ding, Kiran K Solingapuram Sai, Haiying Chen, Kathleen M Hayden, Yongmei Liu, R Nick Bryan, Ilya M Nasrallah, Sudipto Dolui, Mohamad Habes, José A Luchsinger, Robert A Koeppe, Susan R Heckbert, Suzanne Craft
Abstract
Open AccessCardiovascular disease (CVD) risk factors captured in midlife represent potentially modifiable features of CVD, stroke, dementia, and dementia-related neuropathology and are included in dementia risk scores. Subclinical measures of CVD represent degrees of subclinical vascular aging. We hypothesize that subclinical vascular measures associated with dementia may help identify specific structural and functional aspects underlying vascular contributions to cognitive impairment and dementia over and above conventional dementia risk scores. Utilizing a large (n = 1420), racially, ethnically, and regionally diverse group of older adults free from clinical CVD at baseline from the Multi-Ethnic Study of Atherosclerosis (MESA), we assessed relationships between factor composites representing subclinical CVD collected at baseline and neuroimaging biomarkers acquired ∼17-20 years later. Baseline subclinical factors were differentially associated with MRI measures at follow-up. Subclinical composites of arteriosclerosis and atherosclerosis exhibited the strongest associations with brain MRI, including reduced global and hippocampal gray matter volume, greater white matter hyperintensity burden, and to a lesser extent with higher amyloid PET deposition in both unadjusted and adjusted models. Cross-validated prediction models utilizing all subclinical measures demonstrated that baseline subclinical features best predicted MRI measures indexing global and regional gray matter atrophy. Individual measures representing subclinical arteriosclerosis (e.g., small artery elasticity, systemic vascular resistance, carotid distensibility) and subclinical atherosclerosis (e.g., common carotid intimal-media thickness, maximum intimal-media thickness) were most predictive across imaging outcomes assessed. Ultimately, our findings may highlight specific subclinical atherosclerosis and arteriosclerosis vascular pathways consistent with the vascular contributions to cognitive impairment and dementia.