23º SINAPE - Simpósio Nacional de Probabilidade e Estatística

Dados do Trabalho


Título

SURVIVAL ANALYSIS OF THREE ELDERLY COHORTS: BRAZIL, ARGENTINA AND ITALY.

Resumo

The Aim this work wasto estimate the mortality from people aged ≥65 years old in three scenarios with different socio-economic background and urbanization process by using flexible parametric survival models.
The three cohorts coming from Brazil (n=365, 8 years of follow-up), Argentina (n=1800, 10 years of follow-up) and Italy (n=2472, 30 years of follow-up) were considered and only people with ≥65 years old included. Time to death (months) from enrolment and all-causes mortality were considered. Statistical analysis included Frailty Cox’s Model and Flexible Parametric Survival Analysis. Mean (SE) follow-up was 70.8 (2.6) for men and 74.2 (2.0) for women in Brazil, 63.9 (2.1) for men and 67.8 (1.8) for women in Argentina and 298.7 (10.5) for men and 304.9 (11.0) for women in Italy. Frailty Cox model showed significant positive effects of age and an effect modification of High Blood Pressure and Non-communicable diseases but not effects proportionality. Modeling evidenced a positive statistically significant effect of Age, Non-communicable Diseases and Smoking. There was also an effect modification of Non-Communicable Diseases on High Blood Pressure. Overall Flexible Models estimates were more conservative with an increase in their precision. Conclusion: High Blood Pressure and Non-Communicable Diseases are important causal components and strong risk factors of cardiovascular mortality in these countries. The model obtained indicates a common causal model in three cohorts coming from very different environments.

Palavras-chave

Cardiovascular Diseases, Diabetes, Elderly, Hypertension, Survival analysis.

Área

Estatística Aplicada em Ciências Médicas, Saúde e Meio Ambiente

Autores

LICIANA VAZ DE ARRUDA SILVEIRA, Maria del Pilar Diaz, Alberto Rubén Osella