Polygenic Risk Score for Alzheimer's Disease in Caribbean Hispanics

Abstract

Objective Polygenic risk scores (PRSs) assess the individual genetic propensity to a condition by combining sparse information scattered across genetic loci, often displaying small effect sizes. Most PRSs are constructed in European‐ancestry populations, limiting their use in other ethnicities. Here we constructed and validated a PRS for late‐onset Alzheimer's Disease (LOAD) in Caribbean Hispanics (CH). Methods We used a CH discovery (n = 4,312) and independent validation sample (n = 1,850) to construct an ancestry‐specific PRS (“CH‐PRS”) and evaluated its performance alone and with other predictors using the area under curve (AUC) and logistic regression (strength of association with LOAD and statistical significance). We tested if CH‐PRS predicted conversion to LOAD in a subsample with longitudinal data (n = 1,239). We also tested the CH‐PRS in an independent replication CH cohort (n = 200) and brain autopsy cohort (n = 33). Finally, we tested the effect of ancestry on PRS by using European and African American discovery cohorts to construct alternative PRSs (“EUR‐PRS”, “AA‐PRS”). Results The full model (LOAD ~ CH‐PRS + sex + age + APOE‐ɛ4), achieved an AUC = 74% (ORCH‐PRS = 1.51 95%CI = 1.36–1.68), raising to >75% in APOE‐ɛ4 non‐carriers. CH‐PRS alone achieved an AUC = 72% in the autopsy cohort, raising to AUC = 83% in full model. Higher CH‐PRS was significantly associated with clinical LOAD in the replication CH cohort (OR = 1.61, 95%CI = 1.19–2.17) and significantly predicted conversion to LOAD (HR = 1.93, CI = 1.70–2.20) in the longitudinal subsample. EUR‐PRS and AA‐PRS reached lower prediction accuracy (AUC = 58% and 53%, respectively). Interpretation Enriching diversity in genetic studies is critical to provide an effective PRS in profiling LOAD risk across populations. ANN NEUROL 2021;90:366–376

KYNA/Ahr Signaling Suppresses Neural Stem Cell Plasticity and Neurogenesis in Adult Zebrafish Model of Alzheimer’s Disease

“…Significance of the contrast between LOAD and non-LOAD status (LOAD vs. control for Mayo) was performed using empirical Bayes moderation (eBayes function). We meta-analyzed differential expression results from the three TWAS employing a fixed-effect model and using logFC and its standard error (SE) as inputs, consistently with previous studies of transcriptomics in LOAD [36,43,44]. The SE was calculated from the logFC values for each gene, separately in each cohort, and used for standard meta-analysis using the R package "rmeta".…”

Genetic Studies of Parkinson's and Alzheimer's Disease in Latinos/Hispanics: New Insights and Challenges

“…Separately, Sariya et al highlighted differences in the relative contributions of APOE genotype and PRS between CH and non‐Hispanic Whites for AD. They found that the PRS seems to play a bigger role in predicting LOAD in the CH population 6 suggesting that further refinement of PRS in future studies may be useful, though the functional significance of these variances was not addressed in their investigations. A recent study in an African‐American population involving 25 known loci linked with AD in non‐Hispanic Whites revealed that only APOE and six other loci showed significant association.…”

“…Since the heritability of AD and PD for most cases remains unaccounted for and most of the large scale genome wide association studies (GWAS) have focused on subjects with European‐ancestry, 3,4 there is a need for such studies in other ethnicities to identify novel loci/variants and to validate established ones. In this light, two timely studies by Loesch and colleagues and by Sariya and coworkers in this issue attempted to address some of the knowledge gaps in the genetic architecture of AD and PD in Caribbean Hispanics (CH) and Latinos 5,6 …”

“…The AUC was 83% in a small replication autopsy cohort of 33 cases. They also highlighted that a higher CH‐PRS significantly predicted conversion to LOAD using a subset of 1,239 cases who were followed‐up longitudinally 6 . Interestingly, the PRS derived from European or African American ancestry data had lower prediction accuracy.…”

“…Both type I and II errors can occur due to the use of expression data from tissues/cell types that may not be related to the traits and the potential variation of expression levels across different cell types. Improved methodology and mathematical modeling to validate multiple significant associations per locus can improve its accuracy 6 . Genetic variants associated with gene expression or splicing have been shown to modulate functional mechanisms underpinning complex diseases 11 .…”