Discerning neurodegenerative and aging genetic influences on hippocampal subfields trajectories in the Alzheimer’s disease continuum

Abstract

AbstractBackgroundHippocampal atrophy is one of the most characteristic biomarkers of Alzheimer’s disease (AD), and its genetic basis, as well as that of its distinct subfields, is of current interest. However, hippocampal subfields have so far only been assessed largely unrelated to brain atrophy, and the majority of studies were focused on general population samples, explicitly excluding diagnosed patients. In this study, we evaluated whether hippocampal subfields trajectories are a suitable endophenotype to discern between neurodegenerative and aging genetic influences in the Alzheimer’s continuum.MethodThe dataset includes AD patients (N = 113), mild cognitive impairment (N = 426), and control normal (CN) (N = 272) subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset. Individuals were evaluated with 3T MRI at baseline and every 6 months for a total of 96 months. Hippocampal subfields were extracted using longitudinal processing within FreeSurfer(6.0). We used longitudinal mixed‐effect models with random intercept for individuals to investigated the association between hippocampal subfields volumetric changes and genetic predisposition to AD, Parkinson’s disease, frontotemporal dementia (neurodegeneration), as well as intrinsic and extrinsic epigenetic clock acceleration and telomere length (biological aging). Models were adjusted by age, sex, AD‐stage and total hippocampal volume. Disease status‐specific trajectories were also assessed.ResultGenetic predisposition to biological aging (driven by genetic predisposition to telomere length) was significantly associated with subfield changes in presubiculum (β = ‐11.2, p = 4.1e‐04), parasubiculum (β = ‐4.4, padj = 4.4e‐04), molecular layer (β = 5.9, padj = 2.0e‐04), dentate gyrus (DG) (β = 5.79, padj = 4.4e‐04), CA4 (β = 7.2, padj = 4.4e‐05), and HATA (β = ‐1.7, padj = 0.027). Genetic predisposition to neurodegeneration (driven by predisposition to AD) was associated with volumetric changes in CA3 (β = 4.4, p = 0.047). In stratified models, we observed significant volumetric changes due to genetic predisposition to neurodegeneration in controls (CA1, CA4, DG, parasubiculum, molecular layer). In MCI (subiculum regions and CA4) and AD (hippocampal fissure, molecular layer and DG), volumetric changes were mostly associated with genetic predisposition to biological aging processes.ConclusionWe showed differential genetic mechanisms associated with hippocampal subfield trajectories that were specific to Alzheimer’s disease stage. These results may help elucidate new biological mechanisms in the course of the disease as well as discern additional molecular mechanisms related to neurodegeneration and aging.