Self-rated health (SRH) as a predictor of mortality in elderly men living in a medium-size city in Brazil Lı ´via Maria Santiago a, *, Cristiane de Oliveira Novaes a , Ine ˆs Echenique Mattos b a Public Health and Environment, National School of Public Health/Oswaldo Cruz Foundation, Leopoldo Bulho˜es 1480/817 Manguinhos, Rio de Janeiro, RJ 21041-210, Brazil b Department of Epidemiology and Quantitative Methods, National School of Public Health/Oswaldo Cruz Foundation, Leopoldo Bulho˜es 1480/817 Manguinhos, Rio de Janeiro, RJ 21041-210, Brazil 1. Introduction SRH represents the subject’s perception of his own health, based on interpretations of objective physical aspects, mental status, and expectations and comparative references (Hoeymans et al., 1998). SRH has been studied through surveys headed by a single question asking individuals to report their current overall health status, on a scale ranging from excellent to poor (Fayers and Sprangers, 2002; Alves and Rodrigues, 2005). Mossey and Shapiro (1982) conducted in Canada the first epidemiological study that analyzed SRH as a predictive factor for mortality in elderly individuals. In their study, 3128 individuals 65 years or older were followed for 7 years and showed nearly a threefold risk of death among persons that report poor health, as compared to those that rated their health as excellent, in a model adjusted for confounding variables like age, gender, life satisfac- tion, income, and place of residence. The study was followed by others analyzing the associations with SRH in different countries, in study populations with varying socio-demographic and lifestyle characteristics, using distinct methodologies and considering a series of confounding variables. These studies have found SRH to be a good predictor of morbidity and mortality for individuals 60 years or older (Lee, 2000; Kabir et al., 2003; Ishizaki et al., 2006; Lyyra et al., 2006). Some authors evaluated the influence of length of follow-up on the outcome and agreed that the association is stronger during shorter observation periods (Bath, 2003; Benyamini et al., 2003; Deeg and Kriegsman, 2003; Murata et al., 2006; Lyyra et al., 2009). The existence of possible gender differences in the observed association between SRH and mortality has also been analyzed. Some studies have observed such a difference (Grant et al., 1995; Helmer et al., 1999; Benyamini et al., 2000; Bath, 2003; Okamoto et al., 2008; Lyyra et al., 2009), while others have not (Lee, 2000; Brunner, 2006; Min et al., 2006). We found no studies in the literature that analyze such associations in the elderly population in Brazil or in other Latin American countries. Thus, the present study seeks to explore the role of SRH as a predictor of mortality in elderly men living in a medium-size city in Brazil, followed for 2 years. 2. Material and methods 2.1. Data source The data used in this study are from the ‘‘EPOCA’’ Research Project on Population Aging and Cancer. The methodology used for Archives of Gerontology and Geriatrics 51 (2010) e88–e93 ARTICLE INFO Article history: Received 1 September 2009 Received in revised form 5 January 2010 Accepted 7 January 2010 Available online 4 February 2010 Keywords: Self-rated health Elderly men Predictor of mortality ABSTRACT The objective was to examine the role of SRH as a predictor of mortality in elderly men in a medium-size Brazilian city. In 2 years of follow-up, 120 deaths occurred in the study population, with the following main causes: cardiovascular diseases (40%), neoplasms (22.5%), and respiratory diseases (10%). In practically all of the target variable strata, elderly men with fair or poor SRH showed a higher risk of dying as compared to those with excellent or good SRH. In the final model, the variables fair/poor SRH (hazard risk = HR = 1.88, 95% confidence interval = 95%CI = 1.29–2.72), age (HR = 1.05, 95%CI = 1.03– 1.08), public health system as the regular source of care (HR = 1.69, 95%CI = 1.10–2.60), current smoking (HR = 1.94, 95%CI = 1.24–3.04), and acute cardiovascular disease (HR = 1.62, 95%CI = 1.06–2.47) were associated with mortality. We concluded that SRH proved to be a predictive variable for mortality in elderly men after 2 years of follow-up, with nearly a twofold risk of death among men that reported fair or poor health, after adjusting for age, regular use of the public health system, current smoking, and acute cardiovascular disease. Given the importance of poor SRH for predicting mortality in elderly men, health services should incorporate this indicator into health assessments in this population. ß 2010 Elsevier Ireland Ltd. All rights reserved. * Corresponding author. Tel.: +55 21 2598 2727; fax: +55 21 2590 9110. E-mail address: liviamsantiago@gmail.com (L.M. Santiago). Contents lists available at ScienceDirect Archives of Gerontology and Geriatrics journal homepage: www.elsevier.com/locate/archger 0167-4943/$ – see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.archger.2010.01.004