Age adjustment

Used to compare risks for two or more populations at one point in time or for one population at two or more points in time. Age-adjusted rates are computed using the direct method by applying age-specific rates in a population of interest to a standardized age distribution. This eliminates differences in observed rates that result from age differences in population composition. Age-adjusted rates should be viewed as relative indexes rather than actual measures of risk.

Age-adjusted rates are calculated by the direct method, as in:

$$\genfrac{}{}{0px}{}{n}{\genfrac{}{}{0px}{}{\sum }{i=1}}{r}_i(p_i/P)$$

where

r_i = Rate in age group i in the population of interest

p_i = Standard population in age group i

$P=\genfrac{}{}{0px}{}{n}{\genfrac{}{}{0px}{}{\sum }{i=1}}{p}_i$

n = Total number of age groups over the age range of the age-adjusted rate.

Age adjustment by the direct method requires the use of a standard age distribution. The standard for age adjusting death rates and estimates from surveys in Health, United States is the projected year 2000 U.S. standard population. In Health, United States, 2000, the projected year 2000 U.S. standard population replaced the 1970 civilian noninstitutionalized population for age adjusting estimates from most National Center for Health Statistics (NCHS) surveys. Starting with Health, United States, 2001, it was used uniformly and replaced the 1940 U.S. population for age adjusting mortality statistics and the 1980 U.S. resident population that previously had been used for age adjusting estimates from the National Health and Nutrition Examination Survey.

Changing the standard population has implications for racial and ethnic differentials in mortality. For example, the mortality ratio for the Black to White populations is reduced from 1.6 using the 1940 standard to 1.4 using the 2000 standard, reflecting the greater weight that the 2000 standard gives to the older population, in which race differentials in mortality are smaller.

Age-adjusted estimates from any data source presented in Health, United States that use the projected year 2000 U.S. resident population may differ from age-adjusted estimates based on the same data presented in other reports if different age groups are used in the adjustment procedure.

For more information on implementing the 2000 population standard for age adjusting death rates, see “Age Standardization of Death Rates: Implementation of the Year 2000 Standard.” For more information on the derivation of age-adjustment weights for use with NCHS survey data, see “Age Adjustment Using the 2000 Projected U.S. Population.” The projected year 2000 U.S. standard population is available from the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program.

Mortality data

Death rates are age adjusted to the projected year 2000 U.S. standard population (AgeAdj-Table I). Before 2001 data, age-adjusted rates were calculated using standard million proportions based on rounded population numbers (AgeAdj-Table II). Starting with 2001 data, unrounded population numbers are used to adjust age. Adjustment is based on 11 age groups, with one exception. Age-adjusted death rates for Black males and Black females in 1950 are based on nine age groups, with under 1 and 1–4 combined as one group, and 75–84 and 85 and over combined as one group.

National Health and Nutrition Examination Survey (NHANES)

Estimates based on NHANES are generally age adjusted to the projected year 2000 U.S. standard population using five age groups: 20–34, 35–44, 45–54, 55–64, and 65–74, or 65 and older (AgeAdj-Table I). Before Health, United States, 2001, these estimates were age adjusted to the 1980 U.S. resident population.

National Health Interview Survey (NHIS)

Estimates based on NHIS are age adjusted to the projected year 2000 U.S. standard population (AgeAdj-Table I). Before Health, United States, 2000, NHIS estimates were age adjusted to the 1970 U.S. civilian noninstitutionalized population. Information on the age groups used in the age-adjustment procedure is in the footnotes of specific data sets.

AgeAdj-Table I. Projected year 2000 U.S. standard population and age groups used to age adjust data

NOTES: DVS is Division of Vital Statistics. NHIS is National Health Interview Survey. NAMCS is National Ambulatory Medical Care Survey. NHAMCS is National Hospital Ambulatory Medical Care Survey. NHANES is National Health and Nutrition Examination Survey.

SOURCE: National Cancer Institute, Surveillance, Epidemiology, and End Results. Standard populations—single ages.

AgeAdj-Table II. Projected year 2000 U.S. standard population and proportion distribution by age, for age adjusting death rates before 2001

† Figure is rounded up instead of down to force total to 1.0.

SOURCE: Anderson RN, Rosenberg HM. Age standardization of death rates: Implementation of the year 2000 standard. National Vital Statistics Reports; vol 47 no 3. Hyattsville, MD: National Center for Health Statistics. 1998.