ABSTRACT
PURPOSE: To estimate the prevalence of frailty, according to Tilburg Frailty Indicator (TFI) and CHS index, to investigate the correlation between both instruments and to identify the factors associated with this condition in older users of primary health care. METHODS: This is a sectional study with 302 individuals aged 60 years or more of Rio de Janeiro/Brazil. Sociodemographic, health, functional dependence and lifestyle variables were collected. Frailty was evaluated by the TFI (biopsychosocial frailty) and the CHS index (physical frailty). RESULTS: The prevalence of frailty estimated by the TFI was 35.8% and by the CHS index was 23.5%, while 19.2% were considered frail by both instruments. Significant correlations were observed between the two measures (r = 0.675, p < 0.001) and between the CHS index and the physical (r = 0.744, p < 0.001) and psychological domains (r = 0.322, p < 0.001) of the TFI. Considering the TFI, sex, age, healthy lifestyle, osteoarticular diseases, stroke/ischemia and functional dependence in IADL were associated with frailty. By the CHS index, marital status, satisfaction with housing environment, osteoarticular diseases, medication, functional dependence in ADL and IADL were associated with frailty. CONCLUSION: Our study demonstrates that the CHS index and the TFI identify different groups of frail older adults, and the factors associated with physical frailty and biopsychosocial frailty, also differ. Both instruments seem suitable to be used by primary health care professionals in Brazil. Nevertheless, we believe that a fully self-rated assessment is more practical to be applied routinely in this level of attention in the country.