Subjective language difficulties are associated with tau and objective verbal memory in older women with increased AD risk

Background Early cognitive changes in preclinical Alzheimer’s Disease (AD) include memory problems and language difficulties. Subjective cognitive complaints (SCC) are an AD risk factor and are associated with AD pathologies. Women have higher tau levels over men, despite demonstrating a life‐long “verbal memory advantage.” Therefore, SCC may be a stronger indicator of early tau accumulation than objective cognitive changes. Here, we examined the relationships among tau, subjective cognitive function, and objective cognitive function in older women at higher AD risk. Method 16 older women (age: 72.3±3.1years) participated in Women Inflammation and Tau Study (WITS). The inclusion criteria are: 65+ years old, AD polygenic hazard score >50th percentile, and a telephone Montreal Cognitive Assessment score suggestive of mild cognitive impairment (MCI; range 16‐20). Tau was measured with cerebrospinal fluid (CSF) ptau181 and 18F‐MK6240 positron emission tomography (PET) with standardized uptake value ratio (SUVR) calculated in Braak1, correlating with MCI stage. Objective memory and language performance were assessed with the Rey Auditory Verbal Learning Test (RAVLT; immediate and delayed recall) and the Multilingual Naming Test (MINT), respectively. Subjective cognition was measured approximately one year later with the self‐reported Everyday Cognition Scale (ECog) memory and language subscales (higher score indicate more complaints). Spearman partial correlations analyzed the associations among tau, objective cognition, and subjective cognition while controlling for age. Result Greater CSF ptau181 was significantly associated with higher ECog‐Language (rho = 0.62, p = 0.02), higher ECog‐Memory (rho = 0.53, p = 0.51), and lower RAVLT immediate recall (rho = ‐0.54, p = 0.046; Figure 1). More tau in Braak1 was associated with lower RAVLT immediate recall (rho = ‐0.58, p = 0.03) and higher ECog‐Language (rho = 0.45, p = 0.09), but not ECog‐Memory (p = 0.26; Figure 2). Higher ECog‐Language was significantly associated with lower RAVLT immediate (rho = ‐0.59, p = 0.02) and delayed recall (rho = ‐0.69, p0.28). Conclusion Our preliminary results suggest that, in older women with elevated AD risk, language concerns with memory features (e.g., word finding and forgetting object names) may track with early increases in CSF ptau181 and entorhinal cortex tau‐PET. Self‐reported language decline m