The cerebrospinal fluid (CSF) Aβ_42/40 ratio has proven to be a more reliable biomarker for amyloid pathology than CSF Aβ₄₂ in Alzheimer's disease (AD), helping to correctly classify patients with positive tau biomarkers (T+) that would otherwise have remained outside of the AD continuum. It was shown that the Aβ_42/40 ratio better captures a relative decrease of Aβ₄₂ in patients with high CSF Aβ. However, whether patients with high-amyloid (HiA) AD, in whom A+ is defined by the Aβ_42/40 ratio, exactly compare with their low-amyloid (LoA) counterparts, in whom A+ is defined by Aβ₄₂ solely, deserves further analysis.

We retrospectively included patients with A+T+ AD and evidence of cognitive and neurodegenerative changes (N+). LoA patients were operationally defined as patients with T+N+ and low CSF Aβ₄₂, while HiA patients were defined as patients with T+N+ and normal CSF Aβ₄₂ but abnormal Aβ_42/40 ratio. Tau CSF biomarkers, neuropsychological profile, rates of cognitive decline, structural and metabolic imaging, ApoE genotype and brain neuropathology were compared between the HiA and LoA groups.

At the time of the lumbar puncture, LoA patients were significantly younger than the HiA patients (68.9±8.7years vs. 71.8±9.4; P=0.0015) and had a lower Mini-Mental Status Examination (MMSE) (18.7±6.4 vs. 20.7±6.2; P=0.0005). There was no difference in the neuropsychological profile nor in the annual rates of cognitive decline between the two groups with early AD. No differences were retrieved between groups on CSF Tau and P-Tau biomarkers, atrophy and brain metabolism, distribution of the APOE4 allele and APOE4/E4 genotype, and neuropathology.

Overall, our study supports the surrogate use of the Aβ_42/40 ratio as an equivalent to Aβ₄₂ to define AD. We showed that HiA CSF profiles were not associated with differences in cognition, brain structures and metabolism, APOE genotype tau CSF biomarkers or the rates of cognitive decline, but may be the associated with later-onset and early-stage AD.