Tau and Amyloid Plasma Biomarker Analysis in Alzheimer’s Disease Cohorts from Diverse Genetic Ancestries

Abstract

AbstractBackgroundPlasma concentrations of phosphorylated threonine‐181 of Tau (pTau181) and the ratio of amyloid beta isoforms Ab42/Ab40 are biomarkers for differential diagnosis and preclinical detection of Alzheimer disease (AD). However, measurement of these biomarkers is mostly from individuals of non‐Hispanic, European ancestry. Given differences in AD risk, generalizability of these findings is not assured in individuals of diverse ancestry. This study evaluates the utility of plasma pTau181 and Ab42/Ab40 in discriminating clinically diagnosed AD from cognitively intact, age‐matched controls in ancestrally diverse, admixed cohorts.MethodWe measured plasma pTau181 and Aβ42/Aβ40 with Simoa chemistry using the pTau181 AdvantageV2 and NEUROLOGY 3‐PLEX A assays, respectively. Our cohorts consisted of: 642 African Americans (162 AD and 480 cognitively intact (CI)), 906 Puerto Ricans (385 AD and 521 CI), 149 Peruvians (49 AD and 100 CI), 60 Cubans (26 AD and 34 CI), and 246 non‐Hispanic, European ancestry (22 AD and 224 CI). Linear mixed‐effect regression models adjusted for age, sex, population substructure and relatedness followed by Bonferroni correction was applied to identify differences across AD status. Diagnostic performance and construct receiver operator characteristic (ROC) curves were created from logistic regression models.ResultPlasma pTau181 concentrations were increased in individuals with AD compared to CI (p < 2×10−16) taking into account all individuals and in each cohort separately (African Americans, p = 1.2×10−9; Puerto Ricans, p = 7.6×10−9; Peruvians, p = 0.02), and European ancestry, p = 2.2×10−8) except for the Cubans where there was a trend. There was no significant difference in the plasma Aβ42/Aβ40 ratio, however there was a trend towards a decreasing concentration in AD. Using the area under the ROC, pTau181 was more accurate at predicting status than the Aβ42/Aβ40 ratio, but the classification improved when both biomarkers were combined. The accuracy varied widely over the individual cohorts with AUC from 0.845 for African Americans to 0.683 for Peruvians.ConclusionThese results suggest AD biomarkers are generalizable across ancestries, though the predictive value may differ depending on specific ancestral backgrounds. Ultimately, combining genomic and biomarker data from diverse individuals will increase understanding of genetic risk and refine clinical diagnoses in individuals of diverse ancestries.