HAI Book 2025 - Flipbook - Page 730
Cogswell, Petrice
Modeling the temporal progression of plasma and PET AD biomarkers with
cognition
Petrice Cogswell1, Emily Lundt1, Terry Therneau1, Michael Griswold, Nikki Stricker, Mary
Machulda, Jonathan Graff-Radford, Alicia Algeciras-Schimnich, Val Lowe, Christopher
Schwarz, Matthew Senjem, Jeffrey Gunter, David Knopman, Prashanthi Vemuri, Ronald
Petersen, Clifford Jack Jr
1
Mayo Clinic, Rochester, MN, US
University of Mississippi Medical Center, Jackson, MS, US
2
Objective: To aid patient prognosis, it is of interest to better understand associations of Alzheimer9s disease (AD)
biomarker progression with timing of cognitive decline. The goals of this study are to evaluate the association of
the timing of plasma and PET AD biomarker progression with cognitive decline and modifiers of these
associations.
Methods: We included 2369 Mayo Clinic Study of Aging participants with amyloid PET, serial plasma and cognitive
assessments. We fit non-linear mixed effects models with amyloid PET (PiB), tau PET (flortaucipir), p-tau217 (C2N,
and/or Lilly), and memory z-score as outcomes and sex, education, and APOE ε4 carriership as covariates. The
model included a per-participant adjustment (left-right time-shift) for each biomarker indicating how much earlier
or later the biomarker was estimated to progress relative to the population mean, a correlation (R) between
adjustments for each pair of outcomes, and covariate effects on each outcome9s timing. The memory z-score fit
included an up-down down shift to account for baseline variability.
Results: Participants were median age 70 (range 45-91) years, 47% female, with diagnoses of CU (82%), MCI (15%),
and Alzheimer9s dementia (3%). Model fits are shown in Figure 1. All biomarkers9 individual adjustments showed
modest associations with those of memory z-score, R=0.34-0.41 (Figure 2); how early or late an individual
progressed on a biomarker relative to the population mean explained 12-17% of the variation in relative timing of
cognitive decline. APOE ε4 carriership had the strongest covariate effect, e.g. C2N p-tau217 progression 7.5 (5.4,
10.0) years earlier (Figure 3).
Conclusions: Modest associations of the timing of plasma and PET AD biomarkers with memory z-scores indicate
that each provides some information regarding timing of cognitive decline. Stronger associations of APOE ε4
carriership with amyloid PET and plasma p-tau217 than memory is compatible with a more proximal relation with
amyloid pathology.
HAI2025 - 730
