Liver-Specific Polygenic Risk Score Is Associated with Alzheimer's Disease Diagnosis

Background: Our understanding of the pathophysiology underlying Alzheimer's disease (AD) has benefited from genomic analyses, including those that leverage polygenic risk score (PRS) models of disease. The use of functional annotation has been able to improve the power of genomic models. Objective: We sought to leverage genomic functional annotations to build tissue-specific AD PRS models and study their relationship with AD and its biomarkers. Methods: We built 13 tissue-specific AD PRS and studied the scores' relationships with AD diagnosis, cerebrospinal fluid (CSF) amyloid, CSF tau, and other CSF biomarkers in two longitudinal cohort studies of AD. Results: The AD PRS model that was most predictive of AD diagnosis (even without APOE) was the liver AD PRS: n = 1,115; odds ratio = 2.15 (1.67-2.78), p = 3.62x10(-9). The liver AD PRS was also statistically significantly associated with cerebrospinal fluid biomarker evidence of amyloid-beta (A beta(42):A beta(40) ratio, p = 3.53x10(-6)) and the phosphorylated tau:amyloid beta ratio (p = 1.45x10(-5)). Conclusion: These findings provide further evidence of the role of the liver-functional genome in AD and the benefits of incorporating functional annotation into genomic research.

Automated summary

Functional annotation of genomic data can be used to build tissue-specific polygenic risk score (PRS) models for Alzheimer's disease (AD), which are associated with AD diagnosis and cerebrospinal fluid biomarkers. The liver AD PRS model was particularly predictive of AD diagnosis and showed significant associations with amyloid-beta and phosphorylated tau levels.