RESUMO
OBJECTIVES: To study the prevalence of frailty and its relationship to mortality in cohorts born before and after the Second World War using three different frailty measures. METHODS: Cross-sectional data from two cohorts born in 1935 (n=593) and 1945 (n=714) were studied for frailty at the mean age of 70.7 (SD 1.8) years. Frailty was measured using the Frailty Phenotype (FP), the Frail Scale (FS) and the 74-item Frailty Index (FI>0.21 denoted frailty). Information on socioeconomic factors was obtained via a study questionnaire and the data on mortality were obtained from the Population Information System. RESULTS: The prevalence of frailty by FI was more common in the older 1935 cohort than in the 1945 cohort (p<0.001). The percentage of robust subjects was higher in both sexes in the 1945 cohort using both FI and FS. After adjusting for sociodemographic factors, the difference in the prevalence of frailty between the cohorts remained significant in women only (OR 1.9 (95% CI 1.3-2.9), p=0.001). The FI classified people as frail more often (30.2% in the 1935 cohort and 17.5% in the 1945 cohort) than the FS (13.1% and 8.8%) or FP (1.8% and 1.6%). Low financial satisfaction was associated significantly with frailty in both sexes. Low level of education was associated with frailty in women and being unmarried or divorced in men. Frailty was associated to increased mortality using all frailty definitions in the 1935 cohort with a longer follow-up time. CONCLUSIONS: Improved living conditions and health care may have resulted in the lower prevalence of frailty in the 1945 cohort. The present study further strengthens the association between frailty and mortality and poor economic status and frailty. Frailty definitions are in need of further study.