Associations between self and study partner report on the CFI with regional tau in a multi‐cohort study

Abstract

AbstractBackgroundSelf‐report of cognitive decline is an early behavioral manifestation of Alzheimer’s disease at the preclinical stage, often thought to precede study partner‐report. Prior work has shown associations between subjective cognitive decline and amyloid, but less is known about tau deposition, particularly once tau has spread to neocortex. Using a multi‐cohort study of cognitively unimpaired individuals, we examined associations between self‐ and study partner‐report, amyloid and regional tau. Specifically, we hypothesized that greater self‐ and study partner‐report of cognitive decline would be observed with tau in the medial temporal lobe (MTL), whereas study partner‐report would be more strongly associated with neocortical (NEO) tau compared to self‐report.Method698 cognitively unimpaired individuals (55% elevated‐amyloid, mean age = 72.5, 58% female) came from the A4, LEARN, HABS, and SCD/IADL Studies. All participants underwent Flortaucipir‐PET quantified by SUVr. A MTL tau composite and a NEO tau composite were derived. An internally derived amyloid cut‐off was used for each study. Participants and their study partners completed the Cognitive Function Index (CFI), a 15‐item questionnaire asking about cognitive functioning over the last year. A series of linear regression models using tau (MTL vs. NEO) to predict CFI (self vs. study partner) were conducted, adjusting for age, sex, education, amyloid status and cohort.ResultWe first examined amyloid status and CFI and found elevated‐amyloid was associated with higher self‐report (t = 2.8, p = 0.017), but not study partner‐report (t = 1.5, p = 0.12). Building on this model, greater MTL tau was associated with greater self‐report (t = 3.30, p<0.001) and study partner‐report (t = 3.4, p<0.001), but not amyloid status. Greater NEO tau was associated with greater self‐report (t = 3.2, p = 0.002) and study partner‐report (t = 4.1, p<0.00001), but not amyloid status.ConclusionIn this multi‐cohort study of cognitively unimpaired individuals, elevated‐amyloid was associated with greater self‐report CFI, but not study partner‐report. When accounting for amyloid status, both MTL and NEO tau were significantly associated with both self and study‐partner report, but effects were numerically stronger for study partner for NEO tau. These findings suggest that CFI report both from the participants and their study partners is valuable at the preclinical stage, but may evolve with increasing tau spread.