BMI changes from midlife to late life and their relationship with Alzheimer’s disease pathologies

Decreasing body mass index (BMI) from midlife to late life has been linked to increased Alzheimer's disease (AD) risk and cognitive decline; however, the association with in vivo AD pathology remains unclear. This study aimed to clarify the relationship between midlife-to-late-life BMI changes and in vivo AD pathologies, specifically amyloid-β (Aβ), tau, and AD-signature region cerebral glucose metabolism (AD-CM). This observational cohort study included non-demented older adults recruited from the Korean Brain Aging Study for Early Diagnosis and Prediction of Alzheimer’s Disease (KBASE) between January 2014 and December 2020. Participants underwent clinical evaluations and positron emission tomography (PET) imaging to measure brain Aβ, tau, and AD-CM. Midlife BMI was retrospectively estimated from self-reported weight at ages 40–50 years, while late-life BMI was measured during the study. Participants were categorized based on BMI changes as decreasing (≤ − 4%), stable (− 4% to < 4%), or increasing (≥ 4%). Associations between BMI patterns and AD pathologies were analyzed using multiple linear regression models. A total of 268 participants (mean age 66.6 ± 7.9 years; 56.3% women; 22.5% APOE ε4 carriers) were analyzed. Higher midlife BMI correlated significantly with lower AD-CM (β = − 0.009; 95% CI, − 0.015 to − 0.003; p < .01). In contrast, late-life BMI was not significantly associated with AD-CM. Participants with decreasing BMI patterns from midlife to late life exhibited significantly lower AD-CM compared to those with increasing patterns (β = 0.029; 95% CI, 0.003 to 0.055; p = .027). No significant associations were observed between BMI patterns and Aβ or tau deposition. A decreasing BMI trajectory from midlife to late life is associated with enhanced AD-related neurodegeneration, underscoring the clinical importance of monitoring long-term body weight changes in relation to Alzheimer's disease pathophysiology.