A demographic analysis is proposed of mortality patterns in the U.S. elderly population for the 14-year period 1968 to 1981 utilizing race, age, and sex specific mortality data where all medical conditions reported on the death certificate, and not just the underlying cause of death, are analyzed. The purpose of this analysis is to explain recent increases in life expectancy at later ages by examining trends for major chronic diseases over the 14-year period and to determine sources of variation in those trends.
Five specific aims are addressed.
The first aim i s to use life table and categorical data strategies to assess both underlying cause and total mentions trends in chronic disease mortality.
The second aim i s to evaluate sex differentials in multiple cause mortality data for major chronic diseases and to determine if those sex differentials can be related to cohort differentials in cause specific mortality risks.
The third aim i s to examine differences in the complexity of cause of death patterns for the extreme elderly population (i.e., age 85+) and to compare the pattern of conditions reported at those ages, and their changes over time, with those for persons 75-84 (a portion of the """"""""old-old"""""""" population), persons 65-74 (post-retirement), and persons 55-64 (pre-retirement).
The fourth aim i s to determine if there is significant geographical variation in the cross-temporal multiple cause mortality patterns. If significant geographic variation exists, this suggests that changes in mortality patterns occur at different times in different areas (e.g., circulatory disease declines began in certain states significantly before the rest of the country) and that national mortality patterns are really an average of distinct patterns in discrete geographic regions. The fifth aim is to relate the cause specific mortality patterns to underlying health changes based on biologically motivated models of human aging processes and mortality. In the first four aims demographic mortality analyses and Poisson regression strategies will be employed. In the fifth aim, biologically motivated models will be proposed which will extensively employ insights from biomedical theory and auxiliary health data. An important component of this fifth aim is to evaluate the sensitivity of analytic results to various assumptions. The characterization of trends using multiple cause mortality data will improve understanding of mortality patterns and mechanisms in the elderly population, help understand changes in survival patterns in the population and permit changes in mortality and health in the elderly population to be better anticipated.
Showing the most recent 10 out of 54 publications
