Interactive Relations of Arterial Stiffness, Race, and Age to White Matter Integrity

Abstract

Prior research has consistently demonstrated significant associations between greater arterial stiffness and lesser white matter integrity. However, there is a paucity of research examining the relations of arterial stiffness to white matter integrity in socioeconomically diverse samples of urban-dwelling adults. Moreover, to my knowledge, no previous studies have examined the moderating effects of race and age on this association. This study抯 primary aim was to evaluate the interactive relations of arterial stiffness, race, and age to white matter integrity. Arterial stiffness was assessed with pulse wave velocity (PWV) and white matter integrity was measured by diffusion tensor imaging metrics (DTI) fractional anisotropy (FA) and trace (TR). Participants were 166 African American and White urban-dwelling adults (57% White, M age=52.37, SD= 8.94, 30% in poverty, M PWV=7.76) from the Healthy Neighborhoods of Diversity across the Life Span (HANDLS) Scan Study, a neuroimaging sub-study of the larger HANDLS project. Multivariable linear regression examined up to three-way interactions between PWV, self-identified race, and age to FA and TR of the superior longitudinal fasciculus (SLF), corona radiata (CR), corpus callosum, and internal capsule. Nonsignificant three-way interactions were removed from the model to investigate possible two-way interactions. Nonsignificant two-way interactions were then removed to evaluate possible main effects. Analyses revealed nonsignificant interactions and main effects of PWV for six of the eight white matter integrity outcomes. Although two significant two-way interactions were observed between PWV and race to FA in the SLF and CR, conditional effects analyses did not find PWV to be a significant predictor of FA in either tract for African American or White participants. The overall null findings of this study are largely inconsistent with the extant literature and may indicate a true absence of relations among these variables in the present sample. However, several factors, including sample characteristics, the measures utilized, and limited statistical power, may explain the discrepancy between the current results and those of prior studies. Future work can build upon the present study by examining other neuroimaging outcomes, utilizing a larger sample size, and investigating other moderating variables in sociodemographically diverse populations.