Cognitive performance and blood biomarkers: Insights into their relationship and predicting high Amyloid Probability Score in cognitively impaired older adults

Abstract

Background

Understanding associations between cognitive performance and blood biomarkers supports diagnosis and treatment of Alzheimer's disease (AD); however, this relationship remains unclear.

Objective

Exploring associations between the Amyloid Probability Score (APS; PrecivityAD® blood test) and cognitive performance, and between Mini-Mental State Examination (MMSE) and Cogstate Brief Battery (CBB) in cognitively impaired older adults.

Methods

Older adults aged ≥ 60 years with MMSE from 10 to 27, recruited from a cohort in Japan, were assessed with the CBB, and plasma biomarkers to calculate APS. We used Spearman correlation to explore relationships; receiver operating curves and cross-validation to identify the best model for predicting high APS.

Results

Among 46 participants (mean age = 78.3 ± 5.9; mean MMSE = 20.3 ± 4.6), 39.1% did not complete at least one test in the CBB. Lower MMSE were observed in those who were non-completers. MMSE correlated with Detection (Psychomotor function), Identification (Attention), and their composite (r = 0.34−0.52; p < 0.05; 95%CI excluding null value). Higher APS was seen in those who did not complete the One Card Learning (OCL-visual learning and short-term memory). The best model for predicting High APS included OCL completion status, MMSE and Years of Education, achieving accuracy = 0.808 and kappa = 0.606 in participants with MMSE < 21.

Conclusions

Our findings suggest a consideration of incompletion when using CBB in cognitively impaired older adults and a potential approach to differentiate High APS group among those with moderate to severe dementia based on visual learning and short-term memory, global cognition and education level.