Figure 2.

Associations between myo-inositol and both Aβ-PET and tau-PET retention. SUVR, standardized uptake value ratio; mIns/tCr, myo-inositol to total creatine ratio. (A) Co-variation of mIns/tCr and Aβ-PET retention extracted from the same location of the MRS volume. The participants were stratified in APOE ε4 carriers and non-carriers. The translucent area around the regression line represents the 95% confidential interval for the regression estimate. For visualization purposes, the ratio between mIns and tCr was depicted but the statistical analysis was performed on the residualized ratio (see the methods section; Aβ-PET ∼ mIns/tCr: P < 0.001 in APOE ε4 carriers, P > 0.2 in APOE ε4 non-carriers). (B) Results of the voxel-wise analysis in the APOE ε4 carriers: highlighted clusters represent significant (P < 0.05 FWE) positive correlations between Aβ-PET retention and age corrected mIns/tCr from the precuneus region. The colour scale reflects the voxel-wise t-values. Results were projected to surface and overlaid onto MNI (Montreal Neurological Institute) 152 template space using the connectome Workbench (v1.2 https://www.humanconnectome.org/software/connectome-workbench). (C) Co-variation of mIns/tCr and tau-PET retention extracted from the same location of the MRS volume (Fig. 1A). The participants were stratified in APOE ε4 carriers and non-carriers. For visualization purposes, the ratio between mIns and tCr was depicted but the statistical analysis was performed on the residualized ratio (see the methods section; tau-PET ∼ mIns/tCr: P < 0.001 in APOE ε4 carriers, P > 0.3 in APOE ε4 non-carriers). (D) Flow chart representing the mediation analysis in the APOE ε4 carriers. Direct effect = association between Aβ-PET and tau-PET retention extracted from the same region (precuneus/PCC): β = 0.14, P < 0.05; mediation effect of mIns/tCr: β = 0.03, P < 0.05, mIns/tCr explained 17% (β_ratio) of the effect of Aβ-PET on tau-PET retention.