Further Evidence on Recent Trends in the Prevalence and Incidence of Disability Among Older Americans From Two Sources: the LSOA and the NHIS

{{Summary In 1993, Manton, et al. used the National Long Term Care Survey to analyze changes in disability incidence and prevalence in the US during the 1980s. The results of Manton's study were somewhat mixed, necessitating additional analyses. Therefore, Crimmins examines disability change from 1982 through 1993 using the Longitudinal Study of Aging (LSOA) and National Health Interview Survey (NHIS). "Four questions are addressed: (1) is there any change during the 1980s and early 1990s in the prevalence of disability, (2) in the likelihood of older persons becoming disabled, (3) in the likelihood of recovery from disability, or (4) in the likelihood of dying when disabled or nondisabled?" (p. S59) 	Methods: Two datasets are used-the LSOA and NHIS. "The NHIS is a continuous household survey of the noninstitutionalized population of all ages in the United States, which has been ongoing since 1957 with the express purpose of monitoring population health and health-care usage. The annual sample of persons 70 years of age and over averages almost 9,000. . . [The LSOA] began in 1984 as a supplement to the NHIS. Those 70 years of age and over who participated in the 1984 [NHIS] interview became the basis for a longitudinal sample (N=7,527)" (p. S61). LSOA survey intervals are about two years. Disability in the LSOA is measured by ADL and IADL instruments; the NHIS has general disability questions that serve as "shorthand" versions of ADL and IADL instruments. Two logistic models are used to examine changes in disability prevalence over time. One model pools 12 observation dates for the NHIS and 4 dates for the LSOA; the other examines ADL and IADL prevalence separately by using only the LSOA data. Incidence is determined by using discrete-time hazard models on LSOA data. Three transitions are considered; nondisability to disability, disability to nondisability, and both functional statuses to death.
 * title=Further Evidence on Recent Trends in the Prevalence and Incidence of Disability Among Older Americans From Two Sources: the LSOA and the NHIS
 * authors=Crimmins, et al.
 * tags=uw-madison, wisconsin, sociology, demography, prelim, qual, WisconsinDemographyPrelimAugust2009
 * summary===Introduction==

Results
From 1982-1993, visual inspection shows no dramatic trend in disability prevalence in either the NHIS or LSOA. However, statistical analyses of NHIS data reveal very small but statistically significant reductions in disability. Also, statistical analyses of LSOA data reveal that at "each of the three dates after the first interview, the likelihood of having an ADL disability relative to being nondisabled is significantly higher than in 1984" (p. S63). There are no similar differences in IADLs. In terms of incidence and recovery, "The transition rates from 'nondisability to disability' are 16 percent lower in the 1988 to 1990 interval [than the 1984-1986 interval], whereas the transition rates from 'disability to nondisability' are 43 percent higher in this interval. This can be interpreted to mean that the rate of becoming disabled as well as the rate of overcoming disability changed in the direction of improving functioning among the older population" (p. S66). The death rate for either disabled or nondisabled did not change significantly over time.

Conclusions
There is no clear ongoing trend in the prevalence of disability among older Americans over the 1980s and early 1990s. However, the LSOA indicates significant declines in the onset of disability and improvements in recovery rates. Because there are some differences in findings across datasets (e.g., NLTCS data showed a decrease in disability recovery rates from 1984-1990), it is important not to reach a premature conclusion that disability trends are improving. }}
 * pub_date=1997