Objectives: To test a model of frailty by examining factors associated with institutionalization of older people in Canada; to assess whether diagnostic data provided information about risk beyond that provided by data on functional capacity and demographic variables.
Methods: Cross-sectional study of 1258 institutional subjects and 9113 community-dwelling older adults from the Canadian Study of Health and Aging.
Results: Multiple logistic regression analysis showed that female gender, being unmarried, absence of a caregiver, presence of cognitive impairment (including all types of dementia), functional impairment, diabetes mellitus, stroke, and Parkinson's disease were independently associated with being in a long-term care facility.
Conclusion: Frailty appears to be a multidimensional construct, and not simply a synonym for dependence in Activities of Daily Living. Studies of health outcomes in older people should include diagnostic data as well as demographic information and data on functional capacity.