Abstract
Alzheimer's disease (AD) biomarkers may be associated with decline in non-cognitive domains of intrinsic capacity (IC), but such evidence has not been synthesized. This narrative review, based on a structured PubMed search (last search: December 31, 2025), included 119 human studies examining associations of core AD biomarkers (e.g., amyloid-beta (Aβ) and tau protein) and biomarkers of non-specific processes involved in AD pathophysiology (including neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), structural magnetic resonance imaging (MRI), and fluorodeoxyglucose positron emission tomography (FDG-PET)) with composite IC scores and non-cognitive IC domains. Among included studies, 2 investigated composite IC scores, 49 depressive symptoms, 30 locomotion, 29 hearing impairment, 19 vitality, and 1 vision impairment. The very limited longitudinal evidence on composite IC scores suggests that lower IC was associated with increased p-tau181 levels and that higher baseline NfL predicted steeper IC decline, whereas plasma Aβ42/Aβ40 showed no clear association. At the IC domains' level, higher cerebral Aβ deposition was associated with poorer locomotion, especially slower gait, more consistently than other biomarker modalities. Higher levels of tau biomarkers and NfL were more often associated with lower or declining handgrip strength. Depressive symptoms represented the most investigated non-cognitive IC domain, showing consistent longitudinal associations with lower fluid Aβ42, greater cerebral amyloid deposition, and subsequent brain atrophy. Hearing impairment was linked mainly to higher tau and NfL, reduced glucose metabolism, and brain atrophy; evidence for vision impairment was almost totally absent. Overall, the pattern of associations varied by biomarker modality, IC domain, and study design, suggesting that AD-related pathology and neurodegeneration have functional correlates beyond cognition. Methodological quality, formally appraised with the Newcastle-Ottawa Scale and the JBI checklist, was acceptable for most included studies.