A cluster analysis to define human aging phenotypes

Abstract

The definition of a precise and consistent aging phenotype that allows to measure the physical and cognitive decline, as well as the increase of mortality hazard late in life, is a major problem for studies aimed at finding the genetic factors modulating rate and quality of human aging. In this frame, it seems promising the concept of frailty which tends to figure out the subjects who are more vulnerable and more prone to negative outcomes, such as death or hospitalization. Cognitive, functional and psychological measures turned out to be the most effective measures to define frailty, as they condense most of the frailty cycle that occurs in the elderly and is probably responsible of the aging related physical decline. We used MMSE, Hand Grip strength, and GDS as variable parameters in a hierarchical Cluster Analysis (CA) in order to recognise aging phenotypes. By using a sample of 65-85 years old subjects we identified three frailty phenotypes that were consistent from both geriatric and genetic perspectives. Therefore, the method we propose may provide unbiased phenotypes suitable for the identification of genetic variants affecting the quality of aging in this age range. The CA method was less effective in ultranonagenarians, probably due to the high prevalence of frail subjects in this age group that makes difficult to distinguish discrete phenotypes.