Human age-associated traits, such as cognitive decline, can be highly variable across the population, with some individuals exhibiting traits that are not expected at a given chronological age. Here we present differential aging (Δ-aging), an unbiased method that quantifies individual variability in age-associated phenotypes within a tissue of interest, and apply this approach to the analysis of existing transcriptome-wide cerebral cortex gene expression data from several cohorts totaling 1,904 autopsied human brain samples. We subsequently performed a genome-wide association study and identified the TMEM106B and GRN gene loci, previously associated with frontotemporal dementia, as determinants of Δ-aging in the cerebral cortex with genome-wide significance. TMEM106B risk variants are associated with inflammation, neuronal loss, and cognitive deficits, even in the absence of known brain disease, and their impact is highly selective for the frontal cerebral cortex of older individuals (>65 years). The methodological framework we describe can be broadly applied to the analysis of quantitative traits associated with aging or with other parameters.
Keywords: TMEM106B; aging; brain; genome-wide association study; genomics; inflammaging; inflammation; microglia; progranulin.
Copyright © 2017 The Authors. Published by Elsevier Inc. All rights reserved.