Modeling small vessel disease burden and vascular dementia likelihood from NACC data

Area of Science:

• Neuropathology
• Neuroimaging Analysis
• Cognitive Science

Background:

• Cerebrovascular pathology is a significant contributor to dementia.
• Standardized quantification methods for multicenter datasets are lacking.
• Developed harmonized indices for small vessel disease (SVD) and vascular cognitive impairment (VCI) likelihood.

Purpose of the Study:

• To create standardized, harmonized indices for quantifying SVD and VCI.
• To enable reproducible modeling of vascular contributions to dementia in multicenter studies.
• To link neuropathological findings to in-life cognitive outcomes.

Main Methods:

• Six cerebrovascular domains were scored and aggregated into an SVD index.
• A proxy for the vascular cognitive impairment neuropathology guideline (VCING)-Lite was used.
• Principal component analysis and regression modeling were employed to analyze vascular indices against cognitive factors and dementia status.

Main Results:

• SVD scores demonstrated consistency across datasets.
• The SVD-VCING-Lite proxy showed strong correlations (r=0.72, p<0.001) after adjustment.
• Both SVD and VCI indices were associated with poorer global cognition and increased odds of dementia.

Conclusions:

• The developed framework effectively links cerebrovascular pathology to in-life outcomes.
• This approach enables reproducible research on vascular contributions in mixed dementias.
• Harmonized quantification is crucial for multicenter neuropathological studies.