In an updated examination of trends in disability through 2005, Manton et al. (6) concluded that the declining trends in disability observed in the 1990s continued and, in fact, accelerated between 1999 and 2004–2005. During this latter period, they reported that chronic disability, a combined measure of ADL and instrumental ADL limitations, declined 2.2% per annum.

Accurately monitoring changes in disability rates is critical, both in the context of the anticipated doubling of the older population between 2000 and 2030 (7) and because disability is considered “a better predictor of medical and social service need than simple prevalence or incidence figures of disease.” (8) In this study, we examine trends in basic ADL disabilities and functional limitations among older adults from 2000 to 2005 using the American Community Survey (ACS), the largest nationally representative survey of community-dwelling elderly adults (response rates >93%), and the 2004 National Nursing Home Survey (NNHS). By examining basic ADL disabilities and functional limitations separately, we can better examine the differences and similarities in these trends than a combined chronic disease measure allows.

When older adults living in nursing homes were included (see Figures 2A and B), the increasing trends of basic ADL disabilities were dampened slightly. However, there was still a trend toward statistical significance, indicating an increase in the overall rate of basic ADL disabilities among women and both genders combined (p = .053 and p = .052, respectively). No significant changes in basic ADL disabilities for men (p = .108) or in functional limitations for either gender were found (p = .663 for men and p = .296 for women).

Our findings conflict with those of Manton et al. (6) whose analysis of data from the National Long Term Care Survey (NLTCS) found that the percentage of older adults with ADL limitations decreased between 1999 and 2004 from 17.9% to 16.6%. There are three major reasons for the discrepancy between our findings and those of Manton et al. First, the samples vary in their construction. The ACS uses a cross-sectional sample, whereas the NLTCS is composed of an age-adjusted sample that is followed longitudinally, supplemented with a cross-sectional sample of those reaching 65 years between sample waves (6). Because people are followed over time, the Manton et al. sample may have greater dropout rates of people who develop disability, which may underestimate the prevalence rate of disability in the later years. The ACS uses repeated cross-sectional surveys with very high response rates. Second, Manton’s sample included individuals who lived in group quarters such as board and care homes. The analyses shown in our Figures 2A and B excluded those in group quarters other than nursing homes. Residents in these other forms of group quarters need not be disabled. It is unclear whether this exclusion would have biased estimates upward or downward.

Third, the definitions of basic ADL disabilities differ between the two surveys. The NLTCS investigated seven ADLs and defined chronic disability as having a limitation for at least 90 or more days. The ACS included only dressing, bathing, or getting around inside the home in the list of ADLs and defined limitations as those lasting 6 months or more. Furthermore, the NLTCS asked if the respondent was limited in each activity “without help or special equipment”, whereas the ACS just asked if the limitation “made it difficult” to do the task. It is possible that the ACS questions measure more severe disabilities, and these may have a different trajectory than the types of disabilities measured in Manton’s study.

Although the time period 2000–2005 is too short to definitively identify a trend, these data are intriguing. The ACS data provide evidence that the rates of functional limitations in the community did not decline between 2000 and 2005 and that the rates of basic ADL disabilities increased. The inclusion of NNHS data on institutionalized older adults suggests that the much heralded decline of disability rates in the older population during the 1990s may have reached a plateau in the early 21st century.

If our findings are confirmed in future research, several explanatory factors should be considered. First, better survival rates with chronic disease can lead to greater morbidity and basic ADL disability in the very old population. Second, increased rates of obesity among midlife and older adults are translating into increases in both functional and basic ADL limitations (12). Third, improvements in assistive technology and environmental accommodations may be enabling persons with more severe disabilities to remain at home.

The rise in basic ADL disabilities among the community-dwelling elderly adults may be influenced by two demographic transitions. As noted earlier, there was a substantial decline in the number of individuals living in nursing homes, with a 1.67% annual decline between 1999 and 2004. Thus, it may be that individuals with ADL limitations who would have lived in nursing homes a decade or two ago would now live in the community. However, most individuals with ADL limitations also have functional limitations, and therefore, it is harder to interpret the reasons for the leveling in functional limitation rates.

Demographic changes in the older population provide an additional plausible explanation for the rise in ADL disability with the concomitant flattening of functional limitation rates. With older adults living longer and an unusual trough in births during the Great Depression of the 1930s (13), the age distribution of the 65-and-older population is changing quickly. Between 2000 and 2005, the population of adults aged 80 and older rose quickly from 23.5% (14) to 26.2% (15) of the total community-dwelling population aged 65 and older. Basic ADL disability rates are much higher among those in the oldest age bracket. However, the majority of individuals with ADL limitations would have had some functional limitations 5 years earlier.