Fig. 2

AD biomarkers’ influence on awareness (objective vs subjective) among clinical stages. Notes: Similar to Fig. 1, this figure shows projected values of awareness and its relationship to clinical status, amyloidosis, and tauopathy, using a comparison between participant’s objective (logical memory delayed recall score) and subjective (ECog memory score) performances (both being z-transformed based on normal group’s mean and standard deviation). The same codes are used for our variables of interest, i.e., normal participants being presented in the left panels while impaired can be seen on the right; entorhinal cortex tau SUVr values being displayed on the upper line, and inferior temporal cortex on the lower one; and amyloidosis being split in low, medium, and high levels of amyloid burden, represented in yellow, gray, and green, respectively. While we did not observe a significant three-way interaction for the IT model (p = 0.839), we did for the EC model (p = 0.046) in which participants from the clinically impaired group with both high amyloid and tau burden demonstrated a significantly lower awareness (β = −0.25, S.E. = 0.12, p = 0.03, upper left panel, green line) while, at lower levels of amyloid burden, we did not observe this impact. In the normal group, this interaction did not seem to be significant