Using surveys conducted at both the national and local level, numerous estimates of the prevalence of functional dependency among the elderly population have been made. Nagi (1976), using a 1972 probability sample of the continental United States, found that almost 17% of the noninstitutionalized elderly population required assistance with mobility or personal care. Estimates from the 1979 and 1980 National Health Interview Surveys (NHIS) indicate that almost 12% of the noninstitutionalized elderly, or 2.8 million elderly, were dependent in personal care, mobility, household activities, or home administered health care services (Feller 1983; Weissert 1985). Using data from the 1982 National Long-term Care Survey (NLTCS), Macken (1986) reported that 19% or 5 million Medicare enrollees were functionally impaired. Similar estimates were reported by Manton and Soldo (1985) who found 4.6 million disabled elderly using data from the 1982 NLTCS. Dawson, Hendershot and Fulton (1987), using the 1984 NHIS's Supplement on Aging, found that 10% of the elderly population received help performing personal care activities, and almost 22% were receiving help with home management activities. The variations in prevalence estimates by these investigators reflects the wide, variety of definitions, samples and levels of aggregation used by them.
In addition to national estimates, surveys of functional dependency also have been conducted at the subnational level. Notable among these are the Duke Longitudinal Studies of Aging (1955-1976 and 1975-1984), the Manitoba Longitudinal Study on Aging (1970-1977), the Duke OARS Survey (1972-1974), the Massachusetts Health Care Panel Study (1974-1980), the Cleveland OARS General Accounting Office Study (1975-1986), and the Framingham Disability Study (1976-1978).