Low education as a predictor of poor one-year stroke survival in the EMMA Study (Study of Stroke Mortality and Morbidity in Adults), Brazil

Authors


Correspondence: Alessandra C. Goulart, Center of Clinical Research, Hospital Universitário, Av. Lineu Prestes 256, Butantan – Cidade Universitária, CEP 05508-900 São Paulo, SP, Brazil.

E-mail: agoulart@hu.usp.br

Data linking low socioeconomic status and stroke survival in developing countries are scarce [1, 2]. We evaluated formal education and one-year survival in the Stroke Mortality and Morbidity Study [Estudo de Mortalidade e Morbidade do Acidente Vascular Cerebral (EMMA) study] [3]. We prospectively ascertained 430 (53% of men) consecutive first-ever stroke patients (ischaemic and haemorrhagic) in a community hospital in São Paulo, Brazil from April 2006 to December 2008. Data were collected using the STEPS Stroke Manual instructions [4]. Kaplan–Meier method and Cox proportional hazards were used to perform survival analyses according to education (illiterate, 1–7 years, ≥8 years). The mean age in these strata were 75·1, 68·1, and 63·8 years old, respectively. Our stroke survival rate of 74·9% at one-year follow-up was similar to previous studies in both developed and developing countries. Illiterate stroke survivors had the poorest survival rate (60·7%), followed by patients with one to seven years of education (76·4%), and ≥8 years of education (81·1%) (Fig. 1). Particularly for ischaemic stroke, non-education was an independent predictor for fatal events [age-adjusted hazard ratio (HR) = 2·32; 95% confidence interval (CI), 1·26–4·27] that strengthened after adjusting for sociodemographics and cardiovascular factors (multivariate HR = 2·65; 95% CI, 1·37–5·13). Factors associated with a poor survival in a sub-group analysis were to be illiterate in persons <68 years [multivariate HR = 3·67; 95% confidence interval (CI), 1·10–12·45], female gender (multivariate HR = 3·42; 95% CI, 1·10–12·45), living alone (multivariate HR = 2·78; 95% CI, 1·16–6·66), and smoking habit (multivariate HR = 4·27; 95% CI, 1·28–14·19). This study has some strength as a good retention of participants and the use of education to evaluate socioeconomic status avoids the potential contamination of reverse causation. Our limitations include the small numbers of haemorrhagic strokes and lack of information on stroke severity at hospital admission. Concluding, stroke survival was directly related to years of education, especially for ischaemic stroke. Absence of education was a significant marker of fatal stroke in persons under age 68 years, who live alone, and who smoked.

Figure 1.

Kaplan–Meier curves for death after first-ever stroke during one-year according to educational levels.

Ancillary