To determine the frequency of cerebral amyloid angiopathy (CAA), defined by the Boston criteria v2.0, in biomarker-defined Alzheimer’s disease (AD) and to compare demographic, clinical, cognitive, and biomarker profiles between AD patients with probable and nonprobable CAA.

The Boston criteria are widely used in memory clinics to diagnose CAA, yet the prevalence and clinical significance of CAA among cognitively impaired patients with AD remain uncertain.

We retrospectively included 506 participants (mean age 74.0 ± 6.8 years; 58% female) with AD confirmed by amyloid-PET or cerebrospinal fluid (CSF) biomarkers from the Geneva University Hospitals Memory Center. MRI scans were evaluated by a neuroradiologist, independently reviewed by a trained image analyst, and confirmed by an experienced neurologist, the latter two blinded to clinical data. Participants were classified as AD-Probable CAA or AD-Nonprobable CAA (regrouping possible and no CAA) following the Boston criteria v2.0. Group comparisons used chi-square/Fisher tests for categorical variables and t-test/Mann-Whitney tests for continuous variables. Longitudinal cognitive decline was assessed with linear mixed-effects models. Associations between regional cerebral microbleed presence or burden and CSF biomarkers were analyzed using binary and ordinal logistic regression.

Nearly one-third of biomarker-defined AD patients met Boston v2.0 criteria for probable CAA. CAA status was not associated with cognition or decline but reflected higher global amyloid burden, suggesting limited clinical utility of the Boston criteria in memory-clinic AD populations.