We hypothesize that specific cognitive assessments and plasma biomarkers may exhibit heightened sensitivity during the stage of subjective cognitive decline (SCD). The integration of these plasma biomarkers and cognitive assessments could enhance the ability to predict beta-amyloid (Aβ) pathology in individuals with SCD.

A total of 231 participants, including 74 normal controls (NC) and 157 SCD, underwent Aβ and tau PET scans and blood testing for Aβ40, Aβ42, p-tau181, p-tau217, NfL, and GFAP. Cognitive assessments, plasma biomarkers, tau PET SUVr, and demographics were compared between Aβ+ and Aβ− groups within NC and SCD. The least absolute shrinkage and selection operator (LASSO) and logistic regression were employed to perform variable selection and develop predictive models.

We observed significantly worse global cognition, visuospatial memory performance, executive function, and metamemory, as well as higher tau PET SUVr, elevated levels of p-tau 217, p-tau 181, and GFAP, and lower Aβ42/Aβ40 ratios in SCD Aβ+ compared to SCD Aβ-. The model incorporating BVMT-LD and p-tau217 achieved a slightly higher AUC than the model using p-tau217 and Aβ42/Aβ40 (0.94 vs. 0.93). Partial correlation analyses indicated that both auditory verbal memory (AVLT-LD) and visuospatial memory (BVMT-LD) were significantly negatively associated with p-tau217, whereas only AVLT-LD demonstrated a significant negative association with tau pathology severity.

Visuospatial memory deficit and plasma p-tau217 are powerful biomarkers for identifying Aβ+ in SCD. Auditory verbal memory links to tau pathology severity, while visuospatial memory is more sensitive to Aβ deposition, supporting early intervention to prevent AD progression.