Frailty prevalence and associated factors in the Mexican health and aging study: A comparison of the frailty index and the phenotype

Background: Frailty is a relatively new phenomenon described mainly in the older population. There are a number of different tools that aim at categorizing an older adult as frail. Two of the main tools for this purpose are the Fried's frailty phenotype (FFP) and the frailty index (FI). The aim of this report is to determine the prevalence of frailty and associated factors using both FFP and the FI. Methods: Secondary analysis of 1108 individuals aged 60 or older is participating in the third (2012) wave from the Mexican Health and Aging Study (MHAS). The FFP and the FI were constructed and a set of variables from different domains were used to explore associations. Domains included were: socio-demographic, health-related, and psychological factors. Regarding prevalence, concordance was tested with a kappa statistic. To test significant associations when classifying with each of the tools, multiple logistic regression models were fitted. Results: Mean (SD) age was 69.8 (7.6) years, and 54.6% (n= 606) were women. The prevalence of frailty with FFP was 24.9% (n = 276) while with FI 27.5% (n = 305). Kappa statistics for concordance between tools was 0.34 (p < 0.001). Age, years in school, number of past days in bed due to health problems, number of times that consulted a physician last year for health problems, having smoked in the past, and life satisfaction were associated with frailty when using any of the tools. Conclusions: There is a persistent heterogeneity on how frailty is measured that should be addressed in future research. (C) 2016 Elsevier Inc. All rights reserved.