Estimated Pulse Wave Velocity Inversely Associates with Cognition in the Northern Manhattan Study
Botagoz Aimagambetova1, Taylor Ariko3, Hannah Gardener5, Bonnie Levin6, Xiaoyan Sun7, Carolina Gutierrez2, Consuelo Mora-McLaughlin8, Jose Gutierrez8, Mitchell sv Elkind9, Clinton B Wright10, Ralph Sacco11, Digna Cabral2, Tatjana Rundek4
1Clinical and Translational Science Institute, 2Neurology, University of Miami, Miller School of Medicine, 3Evelyn F. McKnight Brain Institute, University of Miami, 4Neurology, Evelyn F. McKnight Brain Institute, University of Miami, 5Neurology, Miller School of Medicine, 6University of Miami, 7University of Miami Miller School of Medicine, 8Columbia University, 9Neurology and Epidemiology, Columbia University, 10Clinical Reseaerch, National Institute of Neurological Disorders and Stroke, 11Neurology, Miller School of Medicine - Univ of Miami
Objective:
Determine whether estimated pulse wave velocity (ePWV) is associated with cognition in a racially and ethnically diverse sample of older adults.
Background:
Arterial stiffness has been associated with cognitive performance. The gold standard for measuring vascular stiffness is pulse wave velocity using arterial tonometry. Recently, ePWV calculated from age and blood pressure has been associated with cerebrovascular disease.
Design/Methods:
We included 1290 stroke-free participants (60% women, mean age 64±8 years, 66% Hispanic, 17% Black, 15% White, mean education 10 years) from the population-based Northern Manhattan Study and calculated ePWV using a non-linear function of age and mean arterial pressure measured at initial visit (ePWV=9.587-(0.402*Age)+(4.560*0.001*Age2)-(2.621*0.00001*Age2*MAP) +(3.176*0.001*Age*MAP)-(1.832*0.01*MAP) where MAP=DBP+0.4*(SBP-DSP). The neuropsychological testing was performed to estimate global cognitive performance and domain-specific episodic memory, executive function, semantic memory, and processing speed. Cognition scores were regressed on ePWV in multivariable linear regression models: Model 1 included age, sex, race/ethnicity, education, time from baseline to cognitive testing; Model 2 added hypercholesterolemia, antihypertensive medication, diabetes, smoking and alcohol use; and Model 3 additionally included white matter hyperintensity volume, silent brain infarcts, and perivascular spaces. Effect modification of ApoE4 was tested by including an interaction term in Model 2.
Results:
Mean ePWV for the total sample was 11±2 and did not differ among sex or race-ethnic groups. ePWV was inversely associated with global cognition (b=-0.035, p<0.05), semantic memory (b=-0.049, p<0.05) and processing speed (b=-0.056, p<0.05) after adjustment for demographics and vascular risk factors. In stratified model, greater ePWV was associated with worse processing speed (b=-0.143, p<0.01) in ApoE4 carriers, but not in ApoE4 non-carriers (b=-0.037, p<0.2).
Conclusions:
Results suggest an important role of arterial stiffening in age-related cognitive function. APOE 4 carriers with age-related systemic vascular stiffening has much worse cognitive performance as compared to that in non-APOE 4 carriers with age-related systemic vascular stiffening.