Characterizing single-shell NODDI-DTI estimates through associations with plasma biomarkers and cognitive performance

White matter (WM) degeneration is an important feature of aging and Alzheimer’s disease (AD). Neurite orientation dispersion and density imaging derived from diffusion tensor imaging (NODDI-DTI) is a technique that allows the estimation of neurite density index (NDI) and orientation dispersion index (ODI) in WM from single-shell diffusion tensor imaging. However, these estimates are tensor-derived approximations rather than direct compartment-specific measures.

We aimed to biologically contextualize NODDI-DTI estimates by examining their associations with plasma biomarkers and follow-up cognition and to assess their correspondence with multi-shell NODDI in an independent cohort. Single-shell diffusion MRI, clinical and plasma biomarker data from 42 older adults were obtained from the Alzheimer’s Disease Neuroimaging Initiative 3 (ADNI3) database. We tested associations between diffusion metrics and plasma biomarkers by performing an exploratory voxel-wise analysis to characterize the spatial distribution and the percentage of statistically significant associations across the WM skeleton and tract-based analyses in a priori AD-vulnerable WM pathways.

We additionally tested whether baseline NODDI-DTI metrics and plasma biomarkers were associated with Montreal Cognitive Assessment (MoCA) performance at a mean 4.8-year follow-up in 25 participants. In a separate ADNI3 cohort (n = 103), tract-wise NDI and ODI estimates obtained using NODDI-DTI and multi-shell NODDI were also compared using partial correlations and agreement analyses. pTau181 was the only biomarker showing widespread voxel-wise associations with diffusion metrics. Higher pTau181 was associated with lower NDI.

In tract-based analyses, higher plasma pTau181 was associated with lower NDI in the fornix and inferior fronto-occipital fasciculus. Higher plasma NfL was associated with lower NDI and higher ODI in the fornix. Higher baseline ODI in the hippocampal cingulum and higher baseline pTau181 were both associated with worse follow-up cognitive performance.

In the independent cohort, corresponding NDI and ODI estimates were correlated across all tracts, although agreement analyses showed negative, tract-dependent mean biases for NODDI-DTI. NODDI-DTI may provide an interpretable framework for summarizing single-shell diffusion variation in relation to AD-relevant biomarkers and cognition, particularly when multi-shell acquisitions are not available. However, its estimates remain closely constrained by the underlying tensor signal and should not be considered independent, compartment-specific or quantitatively interchangeable with multi-shell NODDI measures.

Alzheimer’s Disease Neuroimaging Initiative Accelerated Microstructure Imaging via Convex Optimization Digital Imaging and Communications in Medicine Neurite orientation dispersion and density imaging Threshold-free cluster enhancement with family-wise error correction Data used in preparation of this article were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. The investigators within ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in the analysis or writing of this report. A complete listing of ADNI investigators can be found in the ADNI Acknowledgement List: https://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf.