#113-S life expectancy in a long lived population

Abstract

PURPOSE: To evaluate the changes in age-specific life expectancy and rate of aging from 1960 to 1988 for males and females in the Adventist Mortality Study (AMS) (n = 27517) and the Adventist Health Study (AHS) (n = 34192) combined. The data are also used to model changes in life expectancy and mortality rate in the future using the Lee-Carter model. Differences in life expectancy by subgroups with different life styles are explored. The life styles considered were physical inactivity, diet, smoking status, and educational level. METHODS: We used the abridged life table to calculate life expectancy for all-cause mortality for the ages 30–49 and every five-year interval up to 100 years of age. We used the Lee-Carter method to model life expectancy in the near future. We also used the Multivariate Multiple Decrement Life Table Analysis Program (MMDLP) to calculate the life expectancies after adjusting for confounders. The model is basically non-parametric except for a proportional hazard assumption and the time variable is attained age. We also modeled the life expectancies using linear regression. RESULTS: Subjects with a healthy lifestyle had a higher life expectancy. For all age groups, the life expectancy increased linearly from 1960 to 1988. For males, the life expectancy was 26.04 years for subjects 50-54 years in 1960 versus 30.29 in the year 1988. For males 90–94 years of age the life expectancy was 3.29 years in 1960 versus 3.71 in the year 1988. Mortality rates decreased from 0.00823 in 1960 to 0.00294 in 1988 for males 50–54 year of age and from 0.030 in 1960 to 0.2391 in 1988 for males 90–94 years of age. For females 50–54 years of age, the life expectancy increased from 26.65 in 1960 to 34.34 in the year 1988. For females 90–94 years of ages, it increased from 4.83 in 1960 to 4.95 in 1988. The mortality rates for females 50-54 years of age decreased from 0.00675 in 1960 to 0.00276 in 1988 and from 0.16941 in 1960 to 0.16220 in 1988 for females 90–94 years of age. CONCLUSION: Even in the oldest old (95+ for males and 100+ for females), there was an increase in life expectancy and decrease in mortality rate from 1960 to 1988. The negative intercept on the regression lines at older ages may indicate that the life expectancy has not always increased linearly prior to 1960.