153 Longitudinal tau PET and cognition sensitivity analyses). Results for change in R1 also remained essentially unchanged, except for a large increase in the estimate for the association with decline in MMSE (CI; β= 8.88 SE=4.02 when all cognitive data was included versus β= 15.55 SE=5.54 in sensitivity analyses) (Supplementary Figure 8). Independent effects of tau PET BPND and R1 on cognitive decline Models only including cognitive data that was obtained from the moment of the baseline PET scan onward, yielded highly similar results for baseline tau PET BPND and R1 (Supplementary Figure 9 & 10). Results from models including both annual change in tau PET BPND and change in R1 did show some differences (Supplementary Figure 11 & 12). Associations between tau PET accumulation rate in Braak III-VI was no longer predictive of decline on any of the cognitive tests in CU individuals, whereas in CI individuals the association between tau PET BPND accumulation rate in Braak V/VI became much stronger (β= -3.92 SE=1.93 when all cognitive data was included versus β= -9.25 SE=2.84 in sensitivity analyses). Associations between change in R1 and cognitive decline became stronger (larger estimates) in both CI and CU individuals, but no association remained significant in CU individuals (Supplementary Figure 12). 6
RkJQdWJsaXNoZXIy MjY0ODMw