RT Journal Article SR Electronic T1 P1-139 Frailty impact on older Brazilian survival: 3 years follow-up survey JF Journal of Epidemiology and Community Health JO J Epidemiol Community Health FD BMJ Publishing Group Ltd SP A105 OP A105 DO 10.1136/jech.2011.142976d.32 VO 65 IS Suppl 1 A1 Duarte, Y A1 Corona, L P A1 Lebrão, M L A1 Santos, J L F YR 2011 UL http://jech.bmj.com/content/65/Suppl_1/A105.2.abstract AB Objective In this study we will analyse the survival curve of older Brazilians (≥75 years) according frailty categories (not frail, intermediate, frail) in a large, well-defined sample of older Brazilian in 3 years follow-up.Methods Data comes from a longitudinal survey—SABE Study (Health, Well-being and Ageing) that began in 2000 with a sample (n=2143) older adults (≥60 years) living in São Paulo/Brazil.In 2006 1115 older adults were re-interviewed when the frailty analysis (Fried's model) began. The follow-up were analysed with data from four waves, on 2008 and 2009. Survival analysis was done based on data of 2006 (n=687 olders ≥75 years) up to 2009 (death analyses). Kaplan-Meier Survival Analysis was used to analyse the results considering frailty categories in baseline. Cox proportional hazards model was tested using social demographic and health conditions.Results There were 134 deaths and 116 follow-up losses. In survival analysis, the three strata (not frail, intermediate and frail) did not reach the median survival; only frail elders reached 25% survival time, with 17.2 months, so frail elders presented the worst prognosis. The HR for intermediate is 1.94 and 5.47 for frail, in relation to robust (p trend=0.000). The hazard adjusted model showed a HR 1.6 for pre-frail and 2.9 for frail (p<0.01; p trend=0.003).Conclusions Frailty is associated with mortality in Brazilian elders. Recognition of variables involved in frailty may help to perform active prevention and intervention actions and, consequently, maximise survival.