Abstract Body: Background: Intracranial atherosclerotic disease (ICAD), a common cause of stroke, is associated with cerebral atrophy and cognitive impairment, but the underlying pathophysiology remains unknown. We sought to determine if common age-related neuropathologies modified the association between ICAD and cerebral atrophy.

Methods: This cross-sectional analysis included deceased participants from the National Alzheimer’s Coordinating Center database who had MRI morphometry and autopsy assessments for vascular pathologies and Alzheimer’s disease (AD). We represented ICAD, arteriolosclerosis, and AD dichotomously based on autopsy-determined Circle of Willis atherosclerosis (none-mild vs moderate-severe), arteriolosclerosis (none-mild vs moderate-severe), and National Institute of Aging-Alzheimer’s Association ABC score (none-low vs intermediate-high likelihood of AD), respectively. Our primary outcome of interest was total cortical thickness (mm) on MRI. We conducted adjusted linear regression to determine the association between ICAD and cortical thickness, then tested for multiplicative interaction between arteriolosclerosis and AD substrata.

Results: Among 449 included participants (age at enrollment 77 years [interquartile range 70-83], 45% female, 87% non-Hispanic white), 39% had ICAD, 56% had arteriolosclerosis, and 72% had AD. In fully adjusted models, we found ICAD and AD to be independently associated with cortical thinning (ICAD β-estimate [95% confidence interval, CI] = -2.89 [GJD1] [-5.65, -0.13]; AD β-estimate [95% confidence interval, CI] = -5.61 [-8.44, -2.78]). In interactions models, we found significant associations between ICAD and cortical thickness only in subgroups with coexistent arteriolosclerosis or AD pathology, [GJD2] but we did not detect a significant multiplicative interaction (arteriolosclerosis, p = 0.50, AD, p = 0.35).

Conclusion: In this cross-sectional study, we found ICAD to be independently associated with cortical thinning. Coexistent arteriolosclerosis and AD pathologies did not seem to modify the effect of ICAD on cortical thickness, but we may have been limited by sample size.