Metabolomic Changes Are Predictive of Aging in Laying Hens

Abstract

Abstract Aging in vertebrates is an extremely complex process that is still poorly understood. One confining factor to studying vertebrate aging is the lack of appropriate models. The laying hen is a good model to study vertebrate aging, as it can be maintained under standard housing conditions, its breeds are genetically well defined and it exhibits significant aging phenotypes at around 18 months of age. Furthermore, laying hens are maintained in a challenging realistic environment and possess a fully functional immune system. Here we used, for the first time, metabolomic profiling of laying hens’ blood for identifying biomarkers of aging. Random forest classifier was used to quantify the quality of the markers and found that the markers can predict the correct age group of individuals with 90% accuracy. Animals under time-restricted feeding, a condition known to increase health span, appeared younger under the markers, indicating that the aging biomarkers can also predict the effectiveness of environmental treatments. Additionally, we found that noise, defined as the ratio between the standard deviation and the mean, is an exceptionally robust and universal biomarker of aging, as metabolomic noise increases significantly with age in laying hens, humans, and mice. Our study suggests the laying hen as a useful model to study aging in vertebrates and establishes metabolomic noise as a novel, universal biomarker of aging.