Choosing a standard for adjusted mortality rates

For over half a century, the standard for age-adjustment in mortality studies has been based on the total population according to the 1940 census. The question periodically arises, however, whether a more recent census population might now be more appropriate. Thus, a study using the six censuses from 1940 to 1990 was conducted to see the effect each of these populations would have on the age-adjusted (standardized) death rates. While the size of the age-adjusted rates was affected by the censal standard populations from 1940 to 1990, these populations hardly changed the proportional mortality by age, sex, cause-of-death and geographic area. It appears that a shift from the 1940 standard will not be necessary, although if more detailed comparisons are needed, age-specific death rates can always be used. The 1940 standard also has the advantage of being consistent with many earlier studies.