Abstract
Subtle alterations in awareness may emerge in the preclinical stage of Alzheimer's disease (AD), yet their clinical significance and translational relevance remain unclear. This study aimed to evaluate associations of distinct awareness trajectories with clinical and multimodal AD biomarker measurements in cognitively unimpaired (CU) older adults. This prospective study analyzed data from the Anti-Amyloid Treatment in Asymptomatic Alzheimer's (A4) and Longitudinal Evaluation of Amyloid Risk and Neurodegeneration (LEARN) cohorts (∼4.5-year follow-up). Awareness trajectories were defined using a mixed-effects regression model estimating normative longitudinal changes in the Cognitive Function Index (Participant-Study Partner Discrepancy). Based on individual-specific time slopes, participants were classified into three trajectories: stable awareness, heightened awareness (hypernosognosia), and decreased awareness (anosognosia). Study outcomes included the Preclinical Alzheimer's Cognitive Composite (PACC), Alzheimer's Disease Cooperative Study (ADCS) Activities of Daily Living Prevention Instrument (ADL-PI), Clinical Dementia Rating (CDR), plasma phosphorylated-tau at threonine 217 (p-tau217), Aβ-PET ([ (18) F]-florbetapir), tau-PET ([ (18) F]-flortaucipir), and gray matter volume (GMv) via structural magnetic resonance imaging. The associations of awareness trajectories with clinical and multimodal biomarker measurements were evaluated using the general linear model framework, primarily implemented as mixed-effects, including voxel-wise and Braak-stage regional approaches for neuroimaging data. Sequential-longitudinal multimodal neuroimaging mediation analyses evaluated whether regional tau-PET propagation contributed to the emergence of distinct awareness trajectories through downstream GMv loss. In the full sample (n= 1,643) the mean age was 71.49[±4.72] years, ∼60% female sex, mean education of 16.63[±2.74] years, ∼69% Aβ-PET positive, and ∼27% showing clinical progression on CDR-Global (>0). Compared to stable awareness trajectory (n= 1,325[∼80%]; ∼67% Aβ-PET positive; ∼18% clinical progression), hypernosognosia trajectory (n= 157[∼10%]; ∼68% Aβ-PET positive; ∼36% clinical progression) showed modest clinical implications and limited biomarker associations, including plasma p-tau217, medial temporal tau-PET, and brain structure. In contrast, anosognosia trajectory (n= 161[∼10%]; ∼89% Aβ-PET positive; ∼90% clinical progression) was associated with more adverse outcomes, including steeper cognitive and functional decline, higher risk of progression, greater plasma p-tau217, neocortical tau-PET, and widespread neurodegeneration. Associations between regional tau-PET and awareness trajectories were partially mediated by GMv loss, with sequential Braak-stage II tau-PET effects in hypernosognosia and generalized tau-PET propagation effects extending across Braak-stages II-IV in anosognosia. These findings suggest that distinct awareness trajectories emerge from stage-specific pathological processes, alongside downstream neurodegenerative mechanisms, reflecting separate clinical consequences. This study identifies anosognosia as a high-risk trajectory across the early stages of the AD continuum, while suggesting that hypernosognosia may reflect both age-related and early AD-related processes.