Small Area Estimation of Dependency: Final Report. Correlates of Functional Dependency


In addition to prevalence estimates obtained from population based surveys, researchers have explored the demographic, health status and other factors which typically accompany functional decline. A number of specific correlates of dependency have been suggested in previous work. Among these, increases in physical disability have been significantly associated most often with advanced age (Shanas 1962 and 1968; Jette and Branch 1981; Feller 1983; Branch, Katz, Kniepmann and Papsidero 1984; Manton and Soldo 1985; Palmore, Nowlin and Wang 1985; Weissert 1985; Macken 1986; Dawson, Hendershot and Fulton 1987), and with being female (Shanas 1962 and 1968; Jette and Branch 1981; Branch et al. 1984; Palmore, Nowlin and Wang 1985; Manton and Soldo 1985; Weissert 1985; Manton 1988). However, Feller (1983) found no significant difference in rates of dependency by gender, and Dawson, Hendershot and Fulton (1987) found gender differences to disappear when age structure was taken into account.

Other correlates of decrement in functional ability which have been noted, include being nonwhite (Palmore, Nowlin and Wang 1985; Macken 1986), unmarried or residing with family members (Shanas 1962 and 1968; Palmore, Nowlin and Wang 1985), having a low income (Shanas 1968; Palmore, Nowlin and Wang 1985), and being at the low end of the social class continuum (Shanas 1968)

Most recently Jette and Branch (1985) found living alone to be the strongest correlate of physical disability. While they found advancing age to be related to disability among those who live alone, no relationship between advanced age and disability was found among those who lived with others. They also found men who live with others more likely to report physical disability compared to women, but found no significant gender differences among those who live alone. Among those who lived with others, level of income was inversely related to increasing disability.

In a study of active life expectancy (years free of physical disability), using data from a 1974 Massachusetts health care panel study of noninstitutionalized elderly, Katz et al. (1983) found active life expectancy to decrease with age, and to be shorter for the poor at all ages.

A few researchers have also investigated the relationship of functional dependency to other factors with the use of multivariate methods. Nagi (1976) found physical performance, age, number of conditions, sex, race, emotional performance and health status to explain over 74% of the variation in the dependent variable, independent living. In a longitudinal study using residual analysis Palmore, Nowlin and Wang (1985) found changes in ADL abilities to be predicted by prior ADL abilities, age and physical ratings. Using AID (Automatic Interaction Detection) analysis, Heinemann (1985) found the number of chronic conditions, age, social class, and income to be significant predictors of health decline. Pinsky, Leaverton and Stokes (1987) found younger age and higher education levels to be significant predictors of good functioning among both men and women. Using a split-halves test on a data file created by the merger of the 1977 National Nursing Home Survey and the 1977, 1979, and 1980 National Home Health Survey, Unger and Weissert (1988) found that a model with age and age-squared accurately produced regression-adjusted synthetic estimates of the prevalence of dependency among the noninstitutionalized elderly population.

View full report


"smareaes.pdf" (pdf, 2.82Mb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®