Multidimensional patterns of aging: a cluster-analytic approach

A person-oriented approach, cluster analysis, was used to define patterns of aging in 335 subjects from the Gothenburg study of 70-year-olds (H-70). Five distinct patterns, based on the domains of well-being, physical health, functional capacity, cognitive abilities, and social contacts, were found, and members of the five groups were followed longitudinally over 9 years. Overall, the majority of the subjects were in relatively high performing groups at age 70. Two of the groups had very low levels on all of the domains. Group membership at age 70 was predictive of later performance, suggesting that there is utility in typologies of aging individuals and/or trajectories of aging. This study supports a multidimensional approach to the study of variability in aging. Refining the description of agers, and of aging, can contribute both to basic understanding of aging processes and to practical matters of meeting the long-term needs of elderly people.