CSF polygenic risk AD biomarkers predict brain amyloid and free recall. (December 2021)

• Record Type: Journal Article
• Authors: Kennedy, Gabriel; Moore, Annah M.; Mahoney, Emily R.; Dumitrescu, Logan; Sperling, Reisa A.; Deming, Yuetiva; Cruchaga, Carlos; Hohman, Timothy J.; Mormino, Elizabeth C.
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-12
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.056252
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: The utility of aggregate measures of genetic risk such as polygenic risk scores (PRS) during the decades prior to actual disease onset of Alzheimer's Disease (AD) remains unclear and may allow implementation of early risk trials. Method: A sample of 3006 non‐Hispanic White cognitively healthy participants (Table 1) were selected from the Anti‐Amyloid Treatment in Asymptomatic Alzheimer's (A4) study. Variants were imputed on the TOPMed Imputation Server before entering the PRS calculation pipeline and variants within 1000k base pairs of the APOE locus were removed (chr19:43905796‐45909395). Published GWAS summary statistics (n=3, 146) for cerebrospinal fluid (CSF) amyloid β42, phosphorylated tau (p‐tau181) and total tau (t‐tau) were used to derive PRS. A primary p‐value threshold of 0.01 was chosen after testing the fit of various thresholds (0.01, 0.001, 5e‐8/"GWAS‐sig"). A secondary SNP significance threshold of 0.5 was used for clumped variants calculated with LD clumping (r=0.25, 200kbp). Multiple regression models were used to examine the PRS association with continuous and dichotomous florbetapir amyloid PET and average free recall from the FCSRT, using age, sex, education, and APOE dosage ( e4, e2, coded from 0 to 2 by allele presence) as covariates. P‐values from the main models were corrected for multiple comparisons with the Benjamini‐Hochberg false discovery rate. Result: Polygenic risk of CSF‐Aβ42, p‐tau181, and t‐tau were associated with continuous amyloid burden (p=.008/2.7e‐4/.008) and Aβ42 and p‐tau181 were additionally associated with dichotomous amyloid (p=.023/011) (Table 2) in the expected direction. Each PRS had a significant relationship with free recall (p= .002/.03/.02). After adjusting for amyloid PET, the effects of CSF‐Aβ42 and t‐tau remain for free recall (p=9.8e‐4/.025) (Table 3). Analyses stratified by APOE status revealed stronger associations between polygenic risk and amyloid PET among APOE4 + individuals, whereas stronger associations between polygenic risk and free recall was present among APOE4 ‐ individuals (Table 4). Conclusion: Among preclinical participants, aggregate genetic risk for CSF‐Aβ42, p‐tau181 and t‐tau associated with brain amyloid burden and cognition. Genetic risk testing of CSF biomarkers in non‐AD participants may provide strengthened predictions that associate with potential AD health outcomes independently of APOE .