Effects of CD33 Variants on Neuroimaging Biomarkers in Non-Demented Elders.

Ya-Juan Wang,Yu Wan,Hui-Fu Wang,Chen-Chen Tan,Jie-Qiong Li,Jin-Tai Yu,Lan Tan
DOI: https://doi.org/10.3233/JAD-181062
2019-01-01
Journal of Alzheimer s Disease
Abstract:Two CD33 common variants, rs3826656 and rs3865444, have been identified to be correlated with Alzheimer's disease (AD). Our study examined the effects of the two AD-related CD33 common variants (rs3826656 and rs3865444) on the chosen AD-related brain regions (including hippocampus, amygdala, parahippocampus, middle temporal, entorhinal cortex, and total brain volume) in non-demented elders recruited from the Alzheimer's Disease Neuroimaging Initiative database at baseline and during four-year follow-up. We further tested the effects in an A beta-positive group (including preclinical and prodromal stage of AD) and an A beta-negative group. In the total non-demented elderly population, no associations reached significant levels after FDR correction. In the A beta-positive group, we found that rs3826656 was associated with hippocampal and amygdala volumes (Hippocampus-R: p(c) = 0.0022; Amygdala-L: p(c) = 0.0044; Amygdala-R: p(c) = 0.0066), and rs3865444 was associated with right entorhinal volume (p(c) = 0.0286). The associations of rs3826656 with hippocampal and amygdala volumes in the A beta-positive group were successfully replicated in the prodromal AD group (Hippocampus-R: p(c) = 0.0022; Amygdala-L: p(c) = 0.0022; Amygdala-R: p(c) = 0.0088). These changes became more obvious over time during four-year follow-up. No associations were found between the two CD33 variants and neuroimaging biomarkers in the A beta negative and preclinical AD groups after FDR correction. These results suggested that the two CD33 common variants (rs3826656 and rs3865444) influenced volumes and atrophy rates of AD-related brain regions in non-demented elders. Subgroup analyses showed the effects mainly existed in the A beta-positive group instead of the A beta-negative group, and the effects began in the prodromal AD stage.
What problem does this paper attempt to address?