Cross-sectional relationships between memory scores and neurodegeneration biomarkers in patients with suspected cognitive impairment.

Abstract

Introduction: We aimed to examine whether additional metrics derived from the CERAD Word List Memory Test (WLMT) such as acquisition patterns, error frequencies, and consolidation efficiency, are associated with cerebrospinal fluid (CSF) biomarkers of amyloid and tau pathology, as well as magnetic resonance imaging (MRI) ratings of brain atrophy and small vessel disease, in individuals with suspected cognitive impairment who nonetheless perform within normative ranges on the traditional WLMT measures.

Methods: We included 109 patients from a memory clinic setting (46% female, median age 69.5 years, 55.1% with normal AD biomarkers, 25.7% in the AD continuum) that underwent comprehensive neuropsychological assessment, CSF biomarker analysis, and visual MRI rating. Associations between traditional and additional memory scores and the biomarkers were examined using regression analyses, controlling for demographic variables and adjusting for multiple comparisons.

Results: Correlations were modest and partly mediated by demographics. Traditional scores showed the strongest links, but additional scores "lost access" and "learning ratio" were also associated with lower Aβ CSF levels (ρ = -.23, p = .034) and medial‑temporal lobe atrophy (e.g. learning ratio, ρ = -.30, p = .036), suggesting sensitivity to subtle learning‑consolidation inefficiencies. High burden of white matter hyperintensities was associated with intrusion errors (U = 807.5, p < .05, d' = 0.23) and less stable learning (U = 613.0, p < .01, d' = 0.41).

Conclusions: Both traditional and additional memory scores can assist early detection and risk stratification of neurodegenerative disease in patients with normative performance, but these findings require replication in larger, more heterogeneous cohorts.