Abstract / Summary
Introduction: The adiposity-cognition relationship in aging is controversial. This study aimed to assess its sex- and age-specific associations with Alzheimer's disease (AD) pathology using multimodal biomarkers.
Methods: In two longitudinal cohorts (Hubei Memory and Aging Cohort Study [HMACS], n = 2,349; Alzheimer's Disease Neuroimaging Initiative [ADNI], n = 1,858), we assessed body fat via Clínica Universidad de Navarra-Body Adiposity Estimator (CUN-BAE) and performed sex-/age-stratified analyses using multimodal biomarkers (plasma, cerebrospinal fluid [CSF], Positron Emission Computed Tomography [PET]).
Results: In adults ≥65 years of age, higher baseline and increasing CUN-BAE were associated with lower risk of AD-related mild cognitive impairment (MCI) and dementia, with stratified analyses suggesting distinct patterns by sex and age; however, formal interaction tests were non-significant. Each 1% increase in ΔCUN-BAE at ages 70-74 was associated with lower normal cognition (NC)-to-MCI transition risk by 20%. In females, higher CUN-BAE associated with lower phosphorylated tau 217 (p-tau217) and nonlinearly with amyloid beta (Aβ)42 with an inflection point near 37.6%. In males, increasing adiposity correlated with lower p-tau217 and neurofilament light chain (NfL). Neuroimaging revealed inverse associations with Aβ deposition in females and with tau deposition in Aβ-positive males with MCI.
Discussion: Higher adiposity was associated with lower incidence of AD-related decline and less pronounced pathology; stratified exploratory analyses suggested patterns by age and sex, although formal interaction tests were non-significant. However, these observational findings do not support recommendations of blindly increasing body fat for dementia prevention.