IADLs, the presence of specific health impairments, and the respondents' work histories and disability benefits. The SOA also served as the baseline of a longitudinal study, the Longitudinal Study of Aging (LSOA), intended to study the impact of changes in functional status and living arrangements on institutionalization. The LSOA used three forms of follow-up: the National Death Index was used to locate those SOA respondents who died; Medicare files were used to determine hospital use and costs for respondents who were 65 or older at the time of the SOA interview; and surviving respondents who were 70 or older at the time of the SOA were reinterviewed in 1986 and 1988 (National Center for Health Statistics, 1987b).
Based on the 1988 NHIS, 33.1 million people, or 13.7 percent of the civilian noninstitutionalized population, have some degree of "activity limitation" due to chronic conditions. When one takes into account the estimated 2.2 million people with disabilities who live in institutional facilities such as nursing homes or residential facilities for the mentally retarded or mentally ill (U.S. Department of Health and Human Services, 1989d), the total number of Americans with disabilities is about 35 million. To describe the composition of this population, we begin with data on activity limitation from the NHIS and then use data from other sources to provide different perspectives.
The 33.1 million noninstitutionalized people with activity limitations fall into three groups of roughly equal size. Some 10.3 million (4.3 percent of the population) experience limitations that do not interfere with their major life activities. Another 13.1 million (5.4 percent of the population) are limited in amount or kind of major activities that they can carry out. The remaining 9.7 million (4.0 percent of the population) are unable to carry on the major activity for someone their age (National Center for Health Statistics, 1989a).
The combined prevalence of all three levels of activity limitation increases substantially with age, as shown in Figure 2-1. In 1988, the prevalence of any activity limitation increased from 2.2 percent of children under age 5 to 37.6 percent of adults age 70 or older.
The severity of limitation also increases with age, as Figure 2-1 shows. Among people with activity limitations, an increasing proportion in each age group up to age 70 is unable to carry out their major activities (the most severely affected group). After age 70, however, the proportion unable to carry on their major activities decreases, corresponding to the shift in the