| Popis: |
Introduction: Clinical diagnosis of dementia is difficult due to the presence of multiple etiologies with mixed dementia (MX) the most common. MX has been found at autopsy but has been difficult to diagnose during life. Earlier we reported, a dichotomy clustering method using Alzheimer proteins in CSF and vascular injury from MRI to identify MX patients in a small cohort, that failed to include cognitive tests. We now report, a larger cohort from the ADNI database with cognitive tests added and develop methods for separating the MX group and defining preclinical AD (presence of AD factors with normal cognition) and preclinical vascular disease (VD) subgroups (presence of white matter damage with normal cognition). Methods: We used a clustering approach based on three composite diagnostic axes: 1) AD Factor (ADF) derived from CSF AD proteins (Aβ42 and pTau), 2) Vascular Disease Factor (VDF) calculated from mean free water (mFW) and peak width of skeletonized mean diffusivity (PSMD) in white matter based on diffusion tensor imaging, and 3) Cognition (Cog) based on memory and executive functions. We calculated normalized composite scores with values from 0 to 1 and a cut-off of 0.5 for Cog, ADF, and VDF. We applied this trichotomy method to an ADNI cohort (N=538, mean age = 71.7 years, 48.5% males, and 89.5% with greater than 12 years of education). They were classified as 52.6% normal cognition, and 34.6% MCI by ADNI criteria. Results: Eight biologically defined subgroups were identified which included the MX group with both high ADF and VDF (9.3%) and a preclinical VD group (3.9%), and a preclinical AD group (13.6%). Cog is significantly associated with both ADF and VDF, and the partial-correlation remains significant even when the effect of the other variable is removed (r(Cog, ADF/VDF removed) = 0.46, p |
