Screening the Community-Dwelling Population for Disability

Jack M. Guralnik, Linda P. Fried, Eleanor M. Simonsick, Karen J. Bandeen-Roche, Judith D. Kasper

As described in the Introduction, an age-stratified sample of women age 65 years and older residing in the community in Baltimore, Maryland was evaluated for eligibility for the Women's Health and Aging Study (WHAS). The objective of the screening was to identify and recruit the one-third most disabled women living in the community.

A 20 to 30 minute home interview using computer assisted personal interviewing techniques was administered to women who agreed to participate in the screening assessment. This screener questionnaire included batteries assessing disability status and cognitive functioning, both of which were used to determine study eligibility. Information was also obtained from participants on demographic characteristics, self-reported health status, and history of physician-diagnosed chronic conditions (see Appendix B for the complete screening interview).

All screening interviews were completed personally by respondents. Women who were too cognitively impaired to respond to the screener questions or who were otherwise unable to complete the screening interview themselves were excluded. Participants completing the screening interview were considered eligible for the WHAS if they reported difficulty in one or more items in two, three, or four domains of disability (see Introduction to this monograph) and had a Mini-Mental State Examination (MMSE) score (Folstein et al., 1975) of 18 or higher. A total of 4,137 women agreed to be screened, 296 of whom could not personally complete the screening interview. Thus, the screening interview was administered to 3,841 women.

Disability Patterns in the Screened Population

Four domains of disability were considered in the screening process: upper extremity, mobility, higher functioning tasks required for independent living in the community, and self-care. The top half of Table 1.1 shows the specific items included in each domain and the percentage of women participating in the screening who reported difficulty for these items. For each task, the participant was asked whether, by herself and without help from another person or special equipment, she had any difficulty. Data are also presented on the percentage of women with difficulty in one or more items in a domain.

The highest rate of difficulty was in the mobility domain, with over 49 percent of screened women reporting difficulty in one or more items. About 22 percent of women reported difficulty in the self-care domain. Rates of difficulty for all individual self-care items were slightly higher than national estimates for women age 65 years and older (Dawson et al., 1987). For example, 8.2 percent of the WHAS population and 6.5 percent of a national sample of women age 65 years and older had difficulty with dressing, and 3.8 percent of the WHAS population and 1.7 percent of the national population had difficulty with eating. The WHAS population may have a higher rate of disability than the total U.S. population because it is an urban population with a greater proportion of women with low income and education. Consistent with previous work in this field, the prevalence of difficulty in each domain and for specific items increases dramatically with increasing age (Table 1.1).

The bottom half of Table 1.1 shows the distribution of the number of domains in which a participant had difficulty with one or more items. Women who had difficulty in no domains or only one domain were ineligible for the study and constituted 43.7 percent and 20.1 percent of the screened population, respectively. The percent of women with difficulty in no domains dropped steeply with increasing age, from more than 50 percent in those age 65 to 74 years to slightly more than 22 percent of those age 85 years and older. Most of the women with two domains of difficulty had mobility and upper extremity problems, while the majority of women with difficulty in three domains reported problems with these two domains as well as problems with either higher functioning or self-care. The most severely disabled women-those with difficulty in all four domains-constituted about 13 percent of the screened population, but this percentage rose steeply with increasing age, to 30 percent of those age 85 years and older.

The domain approach used in the WHAS to select a moderately to severely disabled study population identified a cohort with diverse patterns and levels of disability. Many combinations of disabilities are represented in the eligible study population. The domain approach to screening was successful in including women with less common patterns of disability who are typically classified as nondisabled using more conventional approaches to disability assessment, such as screening for mobility or self-care disability alone. For example, a small proportion of women with two domains of disability had problems with upper extremity function and either higher functioning or self-care tasks. This important subgroup of women would have been excluded if mobility difficulty, which was extremely common in those eligible for the study, were a fixed requirement for study eligibility.

Table 1.1 also demonstrates how the domain-based screening approach used here worked well to appropriately exclude certain women who would have been eligible if a single criterion, such as self-care difficulty, had been used. For example, a small number of screened women (0.6 percent) reported difficulty in self-care but in no other domains, and thus they were excluded. Closer examination revealed that these women generally reported only a little or some difficulty in bathing or dressing and had no difficulty in tasks such as walking a quarter of a mile, doing heavy housework, and lifting and carrying 10 pounds. Their disability was thus likely to be mild and it was appropriate for them to be excluded from the study.

Demographic and Health Characteristics

Table 1.2 shows demographic characteristics, self-assessed health status, and MMSE score for the total screened population and according to study eligibility status. A slightly lower percentage of the population was age 65 to 74 years than the U.S. female population in 1990 (53 versus 55 percent) and a slightly lower proportion was 85 years and older (10 versus 12 percent) (U.S. Bureau of the Census, 1992). While Black women made up 8 percent of the U.S. female population age 65 years and older (U.S. Bureau of the Census, 1992), they represented one-quarter of the population screened for the WHAS. The screened population had a broad range of educational attainment: a third had less than 9 years of education and nearly a fifth had more than 12 years. Overall, 42 percent of the screened population in the WHAS had 12 or more years of education compared to 56 percent for women age 65 years and older in the United States (Aging America, 1991). There was also a wide range of income; however, one-fifth did not know their income or refused to provide it.

For each category listed in Table 1.2, the distribution of screener status is shown (columns for eligible, non-disabled ineligible, and cognitively impaired ineligible add to 100 percent). For the total population, 33.9 percent were eligible, 62.6 percent were ineligible because they were disabled in only one or no domains, and 3.5 percent were ineligible because of cognitive impairment (MMSE score less than 18). Eligibility ranged from 28 percent for women age 65 to 74 years to 51 percent for women 85 years and older. Higher eligibility rates were seen for African American women, women with less education, and those with lower income. Married women were younger and had lower eligibility. Less than 12 percent of women reporting excellent health were found eligible for the study. In contrast, 80 percent of women reporting poor health were eligible, and an additional 7 percent of these women were cognitively impaired. The mean MMSE score was slightly higher for the non-disabled ineligible than for those who were eligible. Women classified as cognitively impaired who were administered the test had an MMSE range of 0 to 17 and a mean score of 12.5.

Table 1.3 shows the prevalence of self-reported chronic conditions for the total screened population and according to screener status. For all conditions listed there is a substantially higher prevalence in women eligible for the study than in the non-disabled ineligible. The absolute difference in prevalence rates is greater for relatively common diseases such as myocardial infarction, angina, diabetes, arthritis, and hearing problems. In contrast, the ratio of prevalence rates in the eligible versus non-disabled ineligible is greater for rarer conditions such as congestive heart failure, stroke, and Parkinson's disease. Women excluded from the study because they scored 17 or less on the MMSE had prevalence rates that were similar to the ineligible non-disabled group for most conditions. For stroke, hip fracture, and hearing problems, conditions that have previously been demonstrated to be associated with cognitive impairment, the prevalence was substantially higher in the cognitively impaired group than in the ineligible non-disabled group.

Characteristics of Participants and Nonparticipants Among Women Eligible for the Study

Table 1.4 shows sociodemographic and health characteristics of the 1,409 women who completed the screening interview and were found to be eligible for the study. It also presents this information according to whether women participated in the full baseline evaluation or declined to participate further. Of the 1,409 eligible screener respondents, 1,002 (71.1 percent) participated in the full study. Study participation was defined as completing both the baseline interview and the nurse's examination about 2 weeks later.

Overall, women who participated in the study were very similar to the total eligible population on the characteristics shown in Table 1.4 (first two sets of columns). However, in comparing the eligible participants to nonparticipants (second and third sets of columns), certain differences were seen. Among participants, a larger proportion were age 65 to 74 years and a smaller proportion were age 85 years and older. Blacks participated at a higher rate than Whites, with 28 percent of participants and 20 percent of nonparticipants being Black. The participant group included a somewhat higher proportion with more than 12 years of school, with only slightly lower proportions in the other education subgroups. A substantially higher proportion of nonparticipants did not know or refused to report their income. Marital status was quite similar in those who did and did not participate, and there was little difference in the distribution of self-reported health status among those who did and did not participate in the full study. Mean MMSE score was similar for both groups. In summary, there were no major disparities between eligible women who agreed to participate in the full study and those who declined. The group who entered the study was somewhat younger, more often African American, and more often had greater than a high school education than the group that declined, but the two groups had similar marital status, self-reported health, and cognitive function.

Classification of Disability Categories in This Monograph

As stated above, a screening procedure that assesses multiple domains of function was valuable in selecting a heterogeneous group of moderately to severely disabled women for this study. For the purpose of presenting descriptive data in this monograph, however, a more conventional approach to disability classification is used. Data are presented for study-eligible participants according to three levels of disability: receipt of help from a person to perform one or more basic activities of daily living (ADLs) (bathing, dressing, eating, using the toilet, getting in or out of bed or chairs), no receipt of help but difficulty with one or more ADLs, and moderate disability. The last group includes those who meet the criteria for the study but have no difficulty with ADLs.

This classification system focuses on ADLs because they are the most commonly assessed measure of disability in old age. Clinicians and other care providers, researchers, and policy makers all have experience with these categories of disability and understand the functional problems and general characteristics of older people who have these disabilities. The category termed moderate disability includes those women disabled enough to qualify for the study but not so disabled as to have difficulty with basic self-care activities. It therefore includes women with difficulty in two or three of the domains assessed in the screening interview.

Table 1.5 shows the disability patterns (excluding ADL tasks) for ineligible, non-cognitively impaired women and for women who were eligible and participated in the full study, according to the three categories of disability described above. These data are particularly useful for understanding functional characteristics of eligible women with moderate disability and comparing them with women who were ineligible for the study. Overall, nearly all women classified as moderately disabled had upper extremity problems, virtually all reported difficulty with one or more items in the mobility domain, and almost half had difficulty in the higher functioning domain. Comparing eligible women with moderate disability with women who reported some disability but were ineligible for the study, more than 83 percent of those with moderate disability had difficulty in the upper extremity domain as compared with 22 percent of ineligible women with one domain of disability. More than 99 percent of the moderately disabled women had difficulty in the mobility domain compared with 71 percent of women with difficulty in one domain. Furthermore, more than 60 percent of women with moderate disability had two or more areas of difficulty in mobility, compared with less than one-quarter of women with difficulty in one domain. Finally, there was almost no difficulty in the higher functioning domain in the women with one domain of difficulty, while more than 40 percent of moderately disabled women had problems with this domain. Table 1.5 also shows the total number of tasks for which women reported difficulty; these distributions are substantially different stepping from one domain through the most severe level of disability.

These data clearly demonstrate that women with moderate disability were not as disabled as those with difficulty in ADLs, but they were more disabled than ineligible women who had difficulty in one domain only. The question still remains, however, as to whether there were women with one domain of functional difficulty whom some observers might classify as more disabled than certain study-eligible women with two domains of difficulty. The goal of screening was to have no woman classified as eligible for the study who was less disabled than a noneligible woman, and no woman classified as ineligible who was more disabled than an eligible woman. Using a screening instrument, it is rarely possible to perfectly classify individuals as to any measure of health or disease, and it is likely that a small number of women with mild to moderate disability who should have been included were excluded from this study. However, there is no gold standard by which to measure this. When assessing multiple domains of disability it is sometimes quite subjective as to what combination of disabilities is more severe, and, in fact, which specific health state or pattern of disability individuals would find less desirable may be a matter of personal choice.

Format of Monograph Tables

This chapter describes the population who were screened to obtain a sample of the one-third most disabled women living in the community, the study population for the WHAS. Beginning with Chapter 2, all data presented are limited to the 1,002 women who make up the WHAS study population. Most tables present descriptive information for the total study population and within the three age strata and the three disability groups described above. In general, the tables show the actual number of women evaluated, but all other data are weighted to give estimates for the target population the women in the study represent. Appendix A describes the sampling strategy and gives general variance estimates that may be used to estimate the precision of the population rates and means shown in all tables. For a small number of variables, there was a large age gradient in rates; in these cases the information according to disability status is presented both as unadjusted and age-adjusted rates. For most variables, age adjustment made little difference compared to the unadjusted rates for the disability categories, and only unadjusted rates according to disability status are presented.


Aging America. Trends and Projections. (1991). DHHS Pub. No. (FCoA) 91-28001. Washington, DC: U.S. Department of Health and Human Services.

Dawson D, Hendershot G, Fulton, J. Aging in the eighties: Functional limitations of individuals age 65 years and over. (1987). Advance Data No. 133. DHHS Pub. No. (PHS) 87-1250. Hyattsville, MD: National Center for Health Statistics.

Folstein MF, Folstein SE, McHugh PR. (1975). "Mini-Mental State": A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189-198.U.S.

Bureau of the Census. (1992). Sixty-Five Plus in America. Current Population Reports, Special Studies, P23-178, Washington, DC: U.S. Government Printing Office.

Back to Table of Contents

Help Downloading Files on This Page
Skip Navigational Links
NIA Home  IRP Home     What's New     Contact Us     Accessibility     Disclaimer     Privacy     Site Map     NIA Home         
NIH logo-link to NIH Home Page DHHS logo-link to DHHS Web Site USA.Gov logo-link to USA.Gov Web Site
Updated: Wednesday July 14, 2010