Cross-sectional study of 1,323 non-demented adults (mean age 68.0 ± 8.7 years; 56% female; 25% Aβ-positive) from the pan-European AMYPAD PNHS cohort, comparing four cortical MRI metrics — volume, thickness, gray–white matter contrast (GWC), and mean diffusivity (MD) — against global Aβ burden (Centiloids via PET) and estimated time to Aβ-positivity using regression and change-point analyses.
GWC alterations preceded Aβ-positivity by an average of ~7.7 years (range across ROIs: ~5–8 years), MD changes preceded positivity by ~2.8 years, volume reductions by ~2.0 years, and cortical thinning emerged on average ~0.5 years *after* Aβ-positivity. Higher Aβ burden was most strongly associated with widespread GWC reduction (significant across nearly all cortical regions) and cortical thinning, while MD and volume showed weaker or more focal associations.
- Diffusion MRI was available in only 435 of 1,323 participants, limiting power for MD comparisons. - Multi-site heterogeneous MRI protocols required statistical harmonization (ComBat), which may have introduced unequal measurement variance across metrics. - Change-point estimates carry wide confidence intervals and are cross-sectional; longitudinal validation is needed.
GWC — derived from standard T1-weighted MRI with no extra acquisition — appears to shift years before amyloid PET turns positive, suggesting it could serve as an early, accessible screening or trial-enrichment marker for preclinical AD. Clinicians and trialists should consider adding GWC to structural MRI pipelines alongside conventional thickness measures, as the two capture biologically distinct and complementary processes.
Explore related topics
