Assessment of racial/ethnic representation in Alzheimer’s disease genetics datasets

Abstract

AbstractBackgroundBiased representation of populations in genetic research leads to disparities in genetics knowledge and utility. Here, we compare the racial/ethnic demographics of all participants in nationally funded Alzheimer’s disease (AD) genetics cohorts to demographics of US AD mortality.MethodsWe used death certificate data from the CDC Wide‐ranging Online Data for Epidemiologic Research (WONDER) Database to calculate age‐adjusted AD mortality of adults aged 65 years and older for 2016‐2020. We compared the racial/ethnic distributions in mortality to the racial/ethnic distribution of AD genetics participants across 16 studies in the National Institute on Aging Genetics of Alzheimer’s Disease Data Storage Site (NIAGADS) that met the following filters: Disease = `AD`; Molecular Data = `Genotype`; and Type = `GWAS` as well as five studies in the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium. Demographic data was obtained directly from data or via literature review. In addition to the GWAS data, we also compared racial demographic data from sequencing studies released as part of the Alzheimer’s Disease Sequencing Project (ADSP) Umbrella (dssNIAGADS NG00067).ResultsTable 1 describes the racial/ethnic distribution of the age‐adjusted AD mortality rate of the US population from 2016‐2020. Non‐Hispanic Whites had the highest rate (per 100,000 person‐years): 254.2 (95% CI 253.5‐254.9), and Asian or Pacific Islander (AS/PI) had the lowest rate: 125.6 (123.6‐127.71). The racial/ethnic distribution of the 13,909 AD cases across 21 genotype datasets available in NIAGADS or CHARGE are presented in Table 1. Most (87.6%) of AD cases are non‐Hispanic White, with the least (0.25%) being AS/PI. Preliminary analysis of ADSP data show increased proportions of Hispanic and Black participants (Table 1).ConclusionThe racial/ethnic distribution of AD genetics participants is representative of proportion of deaths, but after adjusting for age and population distribution, AS/PI have 50% lower mortality rate compared to non‐Hispanic Whites while comprising less than 5% of participants in genetics cohorts. While progress is being made to diversify sequencing data, AS/PI and American Indian or Alaska Natives remain underrepresented in AD genetics research, which may lead to inequitable benefits from future research and pharmaceutical treatments.