Higher educational attainment is associated with reduced risk for Alzheimer's disease (AD) dementia, and its protective effect may act through alterations in cerebral blood flow (CBF) that allow for better coping with accumulating neuropathology. Additionally, there are sex differences in both the risk of developing AD as well as the potential protective effects of education. We therefore sought to investigate whether education moderates the association of hippocampal CBF and memory in cognitively unimpaired older adults, and to examine if these interactions were moderated by sex.

Cognitively unimpaired older adults from the Alzheimer's Disease Neuroimaging Initiative (ADNI; 51 men, 50 women) underwent neuropsychological evaluation and arterial spin labeling MRI, which was used to quantify bilateral hippocampal CBF. Sex was defined as sex at birth. Multiple linear regressions assessed (1) the independent associations among education, CBF, and memory performance separately in men and women and (2) the three-way interactions among CBF, sex, and education, followed by sex-stratified analyses. Three outcome measures were examined: Logical Memory Story A immediate and delayed recall, and Rey Auditory Verbal Learning Test (RAVLT) intrusions. All models adjusted for age and APOE epsilon-4 allele frequency, and all models with CBF additionally adjusted for cerebral metabolism (baseline FDG-PET composite) and pulse pressure.

CBF was not associated with education or memory in either women or men. There was a positive association between education and delayed memory in women (ß=0.14, t=2.64, p=0.008) as well as trending, positive associations between education and immediate memory in women (ß=0.09, t=1.79, p=0.074) and education and delayed memory in men (ß=0.09, t=1.94, p=0.054). Three-way interactions among sex, CBF, and education were significant on immediate recall (ß=2.55, t=2.53, p=0.013), delayed recall (ß=2.56, t=2.44, p=0.017), and RAVLT intrusions (ß=-2.28, t=-2.27, p=0.026). In women, there were interactions between education and hippocampal CBF on both immediate (ß=2.49, t=2.90, p=0.006) and delayed recall (ß=2.30, t=2.78, p=0.009), such that as education increased, the strength of the association between CBF and immediate memory increased. There was also an interaction between education and hippocampal CBF on RAVLT intrusions in women (ß=-2.42, t=-3.05, p=0.004), such that as education increased, the strength of the association between CBF and number of intrusions decreased; there was a main effect where in women with lower education, as CBF increased, the number of intrusions increased (ß=0.76, t=2.59, p=0.032); in women with higher education, there was no association between CBF and intrusions. In men, none of these two-way interactions were significant.

These results suggest that, in cognitively unimpaired older women, the relationship between hippocampal CBF and memory is moderated by education level, even when adjusting for several other factors. Specifically, higher education may serve as a protective factor in the hippocampal CBF-memory relationship, and this relationship was sex-dependent, occurring in women only. Further research is needed to examine these relationships longitudinally across the clinical continuum of AD. Additionally, this work needs to be conducted in more diverse samples to allow for analyses investigating the impact of education on the intersection of race/ethnicity and sex/gender.