Author(s): Rockwood K, Howlett SE, MacKnight C, Beattie BL, Bergman H, , Rockwood K, Howlett SE, MacKnight C, Beattie BL, Bergman H, , Rockwood K, Howlett SE, MacKnight C, Beattie BL, Bergman H, , Rockwood K, Howlett SE, MacKnight C, Beattie BL, Bergman H,
Abstract Share this page
Abstract BACKGROUND: Frailty and fitness are important attributes of older persons, but population samples of their prevalence, attributes, and outcomes are limited. METHODS: The authors report data from the community-dwelling sample (n = 9008) of the Canadian Study of Health and Aging, a representative, 5-year prospective cohort study. Fitness and frailty were determined by self-reported exercise and function level and testing of cognition. RESULTS: Among the community-dwelling elderly population, 171 per 1000 were very fit and 12 per 1000 were very frail. Frailty increased with age, so that by age 85 years and older, 44 per 1000 were very frail. The risk for adverse health outcomes increased markedly with frailty: Compared with older adults who exercise, those who were moderately or severely frail had a relative risk for institutionalization of 8.6 (95\% confidence interval, 4.9 to 15.2) and for death of 7.3 (95\% confidence interval, 4.7 to 11.4). These risks persist after adjustments for age, sex, comorbid conditions, and poor self-rated health. At all ages, men reported higher levels of exercise and less frailty compared with women. Decreased fitness and increased frailty were also associated with poor self-ratings of health (42\% in the most frail vs 7\% in the most fit), more comorbid illnesses (6 vs 3), and more social isolation (34\% vs 29\%). CONCLUSIONS: Fitness and frailty form a continuum and predict survival. Exercise influences survival, even in old age. Relative fitness and frailty can be determined quickly in a clinical setting, are potentially useful markers of the risk for adverse health outcomes, and add value to traditional medical assessments that focus on diagnoses.
This article was published in J Gerontol A Biol Sci Med Sci
and referenced in International Journal of Swarm Intelligence and Evolutionary Computation