This study assessed trends in survival to old age and identified the factors associated with longevity among the elderly (age > or = 65 years).
Cohort analysis of demographic surveillance data.
The study was conducted in the Butajira Rural Health Programme Demographic Surveillance Site in Ethiopia. Using data collected between 1987 and 2004, the probability of survival to 65 years and remaining life expectancy for women and men aged 65 years were computed. Cox regression analysis was used to assess survival by different factors.
Although the elderly represented 3% of the population, their person-time contribution increased by 48% over the 18-year period. Less than half reached 65 years of age, with remaining life expectancy at 65 years ranging from 15 years in rural men to 19 years in urban women. Rural residence, illiteracy and widowhood were associated with lower survival adjusted for other factors, whereas gender did not show a significant difference. However, the effect of these factors differed between men and women, as demonstrated by survival curves and Cox regression. Widowhood [hazard ratio (HR) 2.02, 95% confidence interval (CI) 1.59-2.57] and illiteracy (HR 2.26, 95% CI 1.86-2.73) affected males to a greater extent than females, and rural residence was associated with poorer female survival (HR 1.68, 95% CI 1.55-1.83).
The number of elderly people is increasing in Ethiopia, with the chance of survival into older age being similar between men and women and approaching that in developed countries. However, rural women and illiterate women and men, particularly widowers, are disadvantaged in terms of survival.