Abstract
OBJECTIVE: Investigate associations of cognitive and brain reserve with trajectories of memory decline in mid-life and late-life, and whether the relationship of memory decline with atrophy differs as a function of reserve. METHODS: Participants were 989 Dutch middle-aged to older adults from the SMART-MR prospective cohort, followed up to 12 years with up to 3 measurements of memory and brain MRI. Education and Dutch National Adult Reading Test (DART) were used as proxies of cognitive reserve, and intracranial volume (ICV) and baseline brain parenchymal fraction (BPF) for brain reserve. Univariate growth curve models analyzed associations of reserve with memory decline, and multiple-group bivariate growth curve models tested the longitudinal brain-memory relationship as a function of reserve. Models were additionally stratified by mid-life and late-life. RESULTS: Higher DART, education, and BPF were related to a slower rate of memory decline, particularly in late-life, but ICV was not. A positive covariance indicated that an individual who undergoes atrophy also undergoes memory decline-this relationship did not differ across cognitive or brain reserve, but was not present in mid-life. Memory declined slower than brain volume, yet rates were more similar in the low DART, education, and BPF groups. DISCUSSION: Higher cognitive (DART, education) and brain reserve (BPF) work protectively in longitudinal memory change. ICV is an inappropriate proxy of brain reserve, failing to show any association with memory performance at baseline or over time. Deconstructing relationships of reserve capacities with longitudinal cognitive and brain outcomes may identify focus areas with potential for intervention.