Table 2. Prediction of ADAS-Cog13 longitudinal changes.
Analyses adjusted by age, education, and sex. Model with all participants also adjusted for baseline clinical diagnosis. Biomarkers values were classified as normal or pathological to predict cognitive decline. Bonferroni-Holm multiple comparison correction applied.
| All CU Participants (n = 172) | CU Aβ- (n = 100) | CU Aβ+ (n = 72) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Coef. (SE) | p-value | AIC | Coef. (SE) | p-value | AIC | Coef. (SE) | p-value | AIC | |
| Flortaucipir SPARE-Tau index | 0.35 (0.05) | <0.0001 | 3074.3 | 0.16 (0.15) | 0.32 | 980.7 | 0.28 (0.09) | 0.0070 | 713.8 |
| Flortaucipir meta-temporal ROI index | 0.32 (0.05) | <0.0001 | 3094.8 | 0.16 (0.12) | 0.24 | 980.9 | 0.20 (0.1) | 0.071 | 716.9 |
| Flortaucipir average Braak index | 0.26 (0.05) | <0.0001 | 3103.7 | 0.14 (0.12) | 0.31 | 981.5 | 0.18 (0.1) | 0.12 | 716 |
| MRI SPARE-AD | 0.19 (0.05) | 0.0002 | 3129.2 | 0.14 (0.09) | 0.20 | 978.9 | 0.10 (0.11) | 0.36 | 728.7 |
AIC: Akaike information criterion. Coef.: Coefficient; SE: Standard error.