Associations Between Self and Study Partner Report of Cognitive Decline With Regional Tau in a Multicohort Study

Background and Objectives Self-reported cognitive decline is an early behavioral manifestation of Alzheimer disease (AD) at the preclinical stage, often believed to precede concerns reported by a study partner. Previous work shows cross-sectional associations with beta-amyloid (A beta) status and self-reported and study partner-reported cognitive decline, but less is known about their associations with tau deposition, particularly among those with preclinical AD. Methods This cross-sectional study included participants from the Anti-Amyloid Treatment in Asymptomatic AD/Longitudinal Evaluation of Amyloid Risk and Neurodegeneration studies (N = 444) and the Harvard Aging Brain Study and affiliated studies (N = 231), which resulted in a cognitively unimpaired (CU) sample of individuals with both nonelevated (A beta-) and elevated A beta (A beta+). All participants and study partners completed the Cognitive Function Index (CFI). Two regional tau composites were derived by averaging flortaucipir PET uptake in the medial temporal lobe (MTL) and neocortex (NEO). Global A beta PET was measured in Centiloids (CLs) with A beta+ >26 CL. We conducted multiple linear regression analyses to test associations between tau PET and CFI, covarying for amyloid, age, sex, education, and cohort. We also controlled for objective cognitive performance, measured using the Preclinical Alzheimer Cognitive Composite (PACC). Results Across 675 CU participants (age = 72.3 +/- 6.6 years, female = 59%, A beta+ = 60%), greater tau was associated with greater self-CFI (MTL: beta = 0.28 [0.12, 0.44], p < 0.001, and NEO: beta = 0.26 [0.09, 0.42], p = 0.002) and study partner CFI (MTL: beta = 0.28 [0.14, 0.41], p < 0.001, and NEO: beta = 0.31 [0.17, 0.44], p < 0.001). Significant associations between both CFI measures and MTL/NEO tau PET were driven by A beta+. Continuous A beta showed an independent effect on CFI in addition to MTL and NEO tau for both self-CFI and study partner CFI. Self-CFI (beta = 0.01 [0.001, 0.02], p = 0.03), study partner CFI (beta = 0.01 [0.003, 0.02], p = 0.01), and the PACC (beta = -0.02 [-0.03, -0.01], p < 0.001) were independently associated with MTL tau, but for NEO tau, PACC (beta = -0.02 [-0.03, -0.01], p < 0.001) and study partner report (beta = 0.01 [0.004, 0.02], p = 0.002) were associated, but not self-CFI (beta = 0.01 [-0.001, 0.02], p = 0.10). Discussion Both self-report and study partner report showed associations with tau in addition to A beta. Additionally, self-report and study partner report were associated with tau above and beyond performance on a neuropsychological composite. Stratification analyses by A beta status indicate that associations between self-reported and study partner-reported cognitive concerns with regional tau are driven by those at the preclinical stage of AD, suggesting that both are useful to collect on the early AD continuum.