Table 3.
Associations between global network measures and tau-PET
| Network property | Whole sample | Control | Preclinical AD | Prodromal AD | AD dementia |
|---|---|---|---|---|---|
| Network size | − 0.25 (0.03)*** | 0.14 (0.26) | − 0.22 (0.18) | − 0.17 (0.10) | − 0.08 (0.06) |
| Degree | − 0.38 (0.06)*** | − 0.37 (0.46) | − 0.22 (0.31) | − 0.36 (0.17) | − 0.26 (0.10)* |
| Connectivity density | − 0.42 (0.13)** | − 1.03 (0.97) | − 0.04 (0.66) | − 0.48 (0.35) | − 0.42 (0.21) |
| Clustering coefficient | − 0.57 (0.05)*** | − 0.24 (0.38) | − 0.58 (0.26)* | − 0.26 (0.14) | − 0.34 (0.08)*** |
| Path length | − 0.87 (0.08)*** | − 0.25 (0.62) | − 1.04 (0.42)* | − 0.65 (0.23)** | − 0.47 (0.14)** |
| Gamma | − 1.11 (0.09)*** | − 0.23 (0.71) | − 0.79 (0.48) | − 0.65 (0.25)* | − 0.72 (0.16)*** |
| Lambda | − 0.98 (0.09)*** | − 0.29 (0.69) | − 1.15 (0.47)* | − 0.72 (0.25)** | − 0.53 (0.15)** |
| Small world coefficient | − 0.12 (0.10)*** | − 0.21 (0.73) | − 0.70 (0.49) | − 0.63 (0.27)* | − 0.75 (0.16)*** |
Data are presented as β (SE); Network measures are Z trasformed; gamma is normalised clustering; lambda is normalised path length; SUVr values in temporal meta-ROI; Model is adjusted for age, sex, TIV, and connectivity density (clustering, path length, γ, λ, small world-only). *p < 0.05, **p <0.01, ***p < 0.001, FDR corrected.