The association between 73 Alzheimer’s disease risk variants and white matter microstructural decline in aging

Abstract

AbstractBackgroundAlthough several prior diffusion MRI studies have indicated a role of medial temporal lobe (MTL) white matter microstructural decline in Alzheimer’s disease (AD), it is unclear which AD risk genes are associated with this decline. The goal of this study is to leverage multi‐site harmonized diffusion MRI data to investigate the relationship between 73 AD risk variants and microstructural decline in MTL tracts.MethodThe dataset used in this study was collated from three longitudinal cohorts of aging [Alzheimer’s Neuroimaging Initiative (ADNI), Baltimore Longitudinal Study of Aging (BLSA), Vanderbilt Memory & Aging Project (VMAP)]. In total, this dataset included 1,308 participants (70% cognitively unimpaired, 27% MCI, 3% AD) aged 50+ who had both diffusion MRI and genotype data and includes a total of 3,442 imaging sessions (mean number of visits: 2.68 ± 1.66, interval range: 1 ‐ 12 years). Genotype data was preprocessed using a standardized approach, and the dosage of the 73 AD risk variants was extracted. Diffusion MRI data was preprocessed using the PreQual pipeline and free‐water (FW) correction was conducted to obtain FW and FW‐corrected fractional anisotropy (FAFWcorr) maps. FW and FAFWcorr were quantified within 5 MTL tracts, including the cingulum, fornix, uncinate fasciculus, inferior longitudinal fasciculus, which were subsequently harmonized using the Longitudinal ComBat package. Linear regression was conducted to investigate the association between AD risk variants and MTL microstructure at baseline, while linear mixed effects regression was used for longitudinal analysis. All models were corrected for age at baseline and sex and corrected for multiple comparisons using the FDR approach.ResultAt baseline, APOE‐e4 dosage was strongly associated with MTL FW in all tracts, with additional associations for TMEM106B and CR1 dosages with fornix FW (Table 1). Longitudinally, APOE‐e4 dosage was a strong predictor of MTL FW decline in all tracts, and KAT8, CR1, and TMEM106B dosages exhibited significant associations with fornix microstructural decline (Table 2 and Figure 1).ConclusionThis study suggests that although APOE‐e4 is the strongest predictor of white matter microstructural decline, several other AD risk genes (i.e., CR1, SPI1, TMEM106B, KAT8) contribute to decline, particularly in the fornix.