Search

BackgroundAlzheimer’s disease (AD) is a complex multifactorial progressive dementia affecting all human populations. As a collaboration model between the International Hundred K+ Cohorts Consortium (IHCC) and the Davos Alzheimer Collaborative (DAC), our aim was to develop a trans-ethnic genomic informed risk assessment (GIRA) algorithm for AD.MethodsThe GIRA model was created to include a polygenic risk score (PRS) calculated from the AD GWAS loci, theAPOEhaplotypes, and non-genetic covariates including age, sex and first 3 principal components of population substructure. The model was first validated using a ancestrally diverse dataset from the eMERGE network, and subsequently validated in a South-Asian population in the UK and 3 East-Asian populations. The distributions of the PRS scores were also explored in populations from 3 African regions. In two validation sites, the PRS was tested for associated with the levels of plasma proteomics markers.ResultsWe created a trans-ethnic GIRA model for the risk prediction of AD and validated the performance of the GIRA model in different populations. The proteomic study in the participant sites identified proteins related to female infertility and autoimmune thyroiditis and associated with the risk scores of AD, highlighting molecular mechanisms underlying the previously observed correlations between these clinical phenotypes.ConclusionsAs the initial effort by the IHCC to leverage existing large scale datasets in a collaborative setting with DAC, we developed a trans-ethnic GIRA for AD with the potential of identifying individuals at high risk of developing AD for future clinical applications. The PRS scores in this model also contribute new research discoveries for the molecular pathogenesis of AD as demonstrated by the proteomic data.

A control group is defined as a group of people used for comparison. Depending on the type of study, it can be a group of healthy people or a group not exposed to risk factors. It is important to allow researchers to select the appropriate control participants. The Korea Biobank Project-sponsored biobanks are affiliated with the Korea Biobank Network (KBN), for which the National Biobank of Korea plays a central coordinating role among KBN biobanks. KBN organized several working groups to address new challenges and needs in biobanking. The “Normal Healthy Control Working Group” developed standardized criteria for three defined control groups, namely, normal, normal-plus, and disease-specific controls. Based on the consensus on the definition of a normal control, we applied the criteria for normal control participants to retrospective data. The main reason for exclusion from the “Normal-plus” group was blood test results beyond 5% of the reference range, including hypercholesterolemia. Subclassification of samples of normal controls by detailed criteria will help researchers select optimal normal controls for their studies.