The recent decline in mortality of the extreme aged: an analysis of statistical data.

Official U.S. statistics reveal a 26 per cent decline in the mortality rate for the extreme aged population (persons 85 years and over) between 1966 and 1977. This reduction was greater than that exhibited by other adult age groups and substantially larger than earlier declines for the extreme aged. This study examines the possible effect of statistical artifacts, such as errors in population estimation, on the changed mortality trend. Data from the Social Security files on Medicare enrollment are utilized to test the accuracy of population estimates for the extreme aged. Mortality rates are decomposed by selected cause of death, sex, and race to determine the influence of demographic, social, and health factors on the shift in the pattern. The results of the analysis indicate that substantial real reduction in mortality has occurred among the elderly population. Reductions among females continue to exceed those among males, but less markedly than in the past. Rates for nonwhites are subject to particularly large margins of error due to incorrect population estimates. A sharp downturn in the age-adjusted rates for the cardiovascular diseases, which are more common among the aged, appears to account for the relatively greater reduction in mortality in this age group.