Multiscale Organization of Amyloid-Associated Vulnerability Across the Hippocampal Memory System

Alzheimer's disease (AD) selectively affects interconnected components of the hippocampal memory system, yet most MRI biomarkers characterize neurodegeneration using whole-hippocampal or isolated regional measurements. We tested whether amyloid-associated structural vulnerability is selectively concentrated in the cornu ammonis 1 (CA1) and the dentate gyrus (DG), reproducible across independent 1Florida ADRC and ADNI cohorts, and spatially localized within these subfields. We integrated hippocampal subfield volumetry, covariate-adjusted amyloid-associated vulnerability scores, 10,000-resample bootstrap rank-stability analysis, complementary cross-sectional clinical-stage comparisons, exploratory diffusion MRI, and within-subfield occupancy mapping. The cornu ammonis 1 (CA1) region and the dentate gyrus (DG) retained the two highest covariate-adjusted amyloid-associated point rankings in both cohorts. In ADRC, the adjusted scores were 5.06% for CA1 and 4.38% for DG, with bootstrap Top-2 frequencies of 59.02% and 41.44%, respectively; in ADNI, the adjusted scores were 9.10% for CA1 and 9.37% for DG, with Top-2 frequencies of 73.60% and 76.40%, respectively. Bootstrap resampling quantified uncertainty in regional scores and rank positions and supported recurrent top two placement of CA1 and DG, while also demonstrating overlap in uncertainty among several ROIs. Complementary cross-sectional comparisons described CA1 and DG volume differences across clinical stages in ADRC and ADNI. Exploratory diffusion MRI identified effect-size differences in hippocampal-associated pathways, but none of the 20 region-metric comparisons survived Benjamini-Hochberg false-discovery-rate correction. Collectively, these findings support a reproducible cross-sectional pattern of amyloid-associated vulnerability within the hippocampal memory system, centered on CA1 and DG and accompanied by spatially heterogeneous within-subfield structural loss. The diffusion MRI findings are exploratory and provide anatomical context rather than evidence of temporal ordering. Longitudinal studies are required to determine the sequence in which these abnormalities emerge.