| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Roybal Center for Enhancement of Late-Life Function, Sargent College of Health and Rehabilitation Sciences, Boston University, Massachusetts
b School of Public Health, Boston University, Massachusetts
Alan M. Jette, Sargent College of Health and Rehabilitation Sciences, Boston University, Boston, MA 02215 E-mail: ajette{at}bu.edu.
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Applying Nagi's disablement model, we wrote questionnaire items that assessed disability in terms of frequency and limitation in performance of 25 life tasks. We evaluated their validity and test-retest reliability with 150 ethnically and racially diverse adults aged 60 and older who had a range of functional limitations, using factor analysis and Rasch analytic techniques to examine and refine the instrument.
Results. Our analyses resulted in a 16-item disability component with two dimensions, one focused on frequency of performance and the other addressing limitation in performance of life tasks, with two disability domains within each dimension. The frequency dimension consisted of a personal and a social role domain, and the limitation dimension consisted of an instrumental and a management role domain. Expected differences in summary scores of known-functional limitation groups support the validity of this instrument. Test-retest intraclass correlations of the reproducibility of each overall dimension summary score were moderate to high (intraclass correlation coefficients .68–.82).
Conclusions. The Late-Life Function and Disability Instrument has potential to assess meaningful concepts of disability across a wide variety of life tasks with relatively few items.
DESPITE the increased focus on disability as an important research outcome in gerontology, there remains a serious need to improve current instrumentation. Two major problems continue to plague the field of disability measurement: the first is a lack of conceptual clarity regarding what is meant by the term disability, and the second is the limited ability of existing instruments to detect important changes in disability in response to interventions.
Traditionally, disability measures used in gerontology outcome research have been organized along the single construct of activities of daily living (ADLs), with a distinction between the categories of basic and instrumental ADLs (IADLs) (1)(2)(3)(4). More recently, health status instruments have been used that frequently consist of multidimensional scales that contain symptom lists, global health status items, as well as physical, emotional, and social questions that possess no apparent conceptual structure (4)(5). Still other researchers employ performance-based batteries designed to assess the capability to perform specific functional tasks as indicators of functional disability (6)(7). The result is confusion within the gerontology literature with similar terms used to describe outcomes that are operationalized in myriad ways (8). Lack of sensitivity of existing outcome instruments to detect important changes in disability status hinders the evaluation of interventions. Use of existing disability or health status measures not designed for evaluative purposes frequently results in ceiling or floor effects if the content of the measures lacks sufficient breadth or if increments of ratings are too global (9).
The Late-Life Function and Disability Instrument (Late-Life FDI) was designed to assess and be responsive to meaningful change in two distinct outcomes: function and disability. We employed Nagi's disablement framework (10)(11) as the underlying conceptual scheme for assessing the functional limitations and disability components of Nagi's model. Functional limitations pertain to limitations in the person's ability to do discrete actions or activities. Disability refers to the person's performance of socially defined life tasks expected of an individual within a typical sociocultural and physical environment. In this article, we describe the construction of the disability component of the Late-Life FDI and the results of a field test of the measure's reliability and validity. The function component of the Late-Life FDI is described elsewhere (12).
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
We hypothesized that it would be important to assess the frequency of performing life tasks and the person's limitation in his or her capability to perform each task. We hypothesized that interventions could differentially affect frequency and/or limitation in performing life tasks. Accordingly, the frequency dimension describes the individual's regularity of participating in life tasks. Frequency questions are phrased, "how often do you do a particular task?" with response options (and prompts) of "very often" (a major part of your life), "often" (a regular part of your life), "once in a while" (an occasional part of your life), "almost never" (rarely a part of your life), and "never." The limitation dimension describes capability of performing these life tasks. We defined limitation to include both personal (health, physical, or mental energy) and environmental (transportation, accessibility, or socioeconomic) factors. Limitation questions ask "to what extent to you feel limited in doing a particular task?" with response choices of "not at all," "a little," "somewhat," "a lot," and "completely." We framed the disability questions in a general fashion because we were interested in including factors other than health conditions (e.g., physical environment) that might limit disability outcomes.
Sampling Procedures
We recruited by telephone a convenience sample of community-dwelling adults who were 60 years of age and older. Exclusion criteria included (i) moderate or severe cognitive impairment, (ii) hospitalization of more than 1 night within the past 6 months, or (iii) inability to lift oneself out of bed. Cognitive status was defined from scores on the Short Portable Mental Status Questionnaire (SPMSQ) (13). Errors in five (or more) out of 10 items constituted moderate to severe intellectual impairment. Scores from the 10 physical function items (PF-10) of the Short Form-36 Health Survey (SF-36) (14) were used to classify subjects into four physical functioning strata based on cut-scores designed to distribute individuals into distinct ability levels. This has been used as screening criteria in previous work (15)(16)(17). Eligible subjects were interviewed by a trained interviewer in their homes. The two components of the Late-Life FDI took, on average, 25 minutes to administer. A random subset of 15 subjects completed a second interview within 1 to 3 weeks after the initial interview to assess test-retest reliability of the disability component.
Exploratory Factor Analysis
A series of factor analyses was used to identify latent factors that could be responsible for the covariation in the disability data. The principal axis method was used for the initial extraction of factors with orthogonal rotation. A one-factor and a two-factor model were tested with comparisons of percentage of the variance explained by each factor in the two models. A maximum likelihood (ML) chi-square test was performed on the one-factor model. Cronbach alpha values were calculated for the sets of items from the one-factor and the two-factor models to confirm that the item-compositions of the retained factors were correlated (i.e., measure the same construct).
Rasch Analysis
Using WINSTEPS (18), a series of one-parameter Rasch rating scale analyses was performed to estimate item calibrations along a common disability scale (19). Separate solutions were derived for each of the two disability dimensions, frequency and limitation, as well as for subsequent domains. The Rasch rating scale model used in these analyses assumed that each disability dimension (frequency and limitation) had a similar underlying structure within its own common 1–5-point rating scale. Thus, tasks that people did not perform frequently would end up on one end of the continuum, whereas tasks that most people performed frequently would end up on the opposite end of the continuum. Similarly, tasks on which many people reported severe limitations (i.e., "extremely limited—cannot do") and tasks on which most people reported few or no limitations would end up on opposite ends of the continuum. Item calibrations were expressed in log-odd units (logits) positioned along an interval scale (20). For ease of interpretation, we transformed this raw logit scale to a 0–100 scale. Item calibrations and person scores are aligned along the same 0–100 continuum. Person scores are based on conversion of raw score counts to each transformed 0–100 scale (21). By our convention, items that were not frequently performed and items in which most people expressed limitations were associated with scores of increasing magnitude, with scores approaching 0 indicating poor capability and infrequent performance, and scores approaching 100 characterizing good capability and frequent performance.
Dimensionality of a scale was tested by demonstrating invariance of estimated item and person measures with the help of goodness-of-fit tests. An item that misfit the model indicated the item failed to discriminate between high and low performers in a way consistent with other items on the scale. Evidence of misfit suggested that the item does not belong to that particular dimension, or the item may be ambiguous in its interpretation by respondents so that it adds little to the scaling of capability or performance. We reported the information-weighted fit statistic (infit), which is standardized to approximate a mean of zero and a standard deviation of one. The infit statistic of 2.00 was used as the criterion for deciding if an item fit statistic was unexpected according to the predicted model. This value was used because the infit statistic approximately follows a t-distribution, and a value of 2.00 has about a .05 probability of occurrence.
Reliability and Validity Analysis
The short-term stability of the disability component of the Late-Life FDI was assessed by test-retest reliability intraclass correlation coefficients (22) between initial and follow-up scores provided 1 to 3 weeks later (mean = 12 days; SD 8.3). As a measure of validity, we hypothesized that severity of disability (as determined by the Late-Life FDI disability component) would be positively associated with severity of functional limitation (based on SF-36 physical functioning scale, PF-10). For each scale and dimension, multiple range tests [least significant difference test (LSD)] were performed as well as post-hoc analyses using an adjusted alpha level based on the three pair-wise comparisons of interest (Bonferroni). Using the adjusted alpha, differences in mean estimates were tested for statistical significance.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
2 = 218.28, p < .0001). Although four potential factors were identified by the minimum eigenvalue of 1.0, the two-factor solution was chosen, as it explained 39.6% of the variance. We labeled the first factor a social role domain because it reflected the frequency of performing various social and community tasks. A second factor was labeled as a personal role domain because it reflected the frequency of performing various personal tasks. One item ("take part in own exercise program") had low factor loadings in both the one- and two-factor solutions, but was retained because of the importance of the item to the overall disability concept. The correlation between the social role and the personal role domains based on factor analysis was r = .432 (p < .01).
|
2 = 201.39, p < .0001). Two factors were identified by eigenvalues greater than 1.0. This two-factor model explained 53.9% of the variance. The composition of this dimension differed in comparison with the two-factor solution for the frequency dimension. Twelve items were loaded on a factor we labeled instrumental role domain that encompassed activities at home and in the community. A second factor emerged that we labeled management role domain because it appeared to identify items that involve organization or management of social tasks that involve minimal mobility or physical activity. The correlation between the instrumental role and management role domains based on factor analysis was r = .567 (p < .01). Fig. 1 displays the relation of each questionnaire item to the frequency and limitation dimensions and their disability domains.
Disability Instrument Scaling
Fig. 2 and Fig. 3 depict the hierarchical orders of items on the frequency and limitation dimensions of the Late-Life FDI disability component. Items are arranged according to item calibrations indicating a continuum of frequency of participation (Fig. 2) or degree of limitation (Fig. 3). Item separation statistics for the frequency dimension (item separation = 9.39) and its domains [social role (item separation = 8.67) and personal role (item separation = 6.35)] as well as the limitation dimension (item separation = 7.39) and its domains [instrumental role (item separation = 7.48) and management role (item separation = 5.88)] all exceed a value of 3.0, an indicator of good spread of items along a scale.
|
|
|
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Our analyses revealed two role domains within the frequency and limitation dimensions. Within the frequency dimension, domains of social role and personal role were identified; within the limitation dimension, domains of instrumental role and management role emerged. These role domains are only moderately correlated with each other, thus supporting the generation of distinct domain scores within the frequency and limitation dimensions of the disability component. Rasch analyses confirmed the hierarchical nature of each disability dimension and role domain, confirmed the adequate spacing of items along each dimension, and confirmed that Rasch-derived mean summary scores distinguish known-disability subgroups.
We were not surprised to see differences in the item grouping of the role domains within each disability dimension. Previous studies have found that responses to self-report questions differ depending upon whether older persons responded to what they actually do versus what they are capable of doing (27)(28). In our sample, the grouping of items in the dimension of frequency seemed to be independent of mobility and physical skills, as items that require competence in physical functioning loaded on different factors ["take care of local errands" (personal) and "go out with others" (social)]. Instead, items converged around domains of personal role and social role. In contrast, responses to questions about limitations in performing tasks appear to reflect the importance of physical skills, as a large factor (instrumental role) emerged that seems to represent the ability to move around the home and community. The second factor in the limitation dimension grouped items around activities involving management skills, such as communication and social planning. This management role, linked to cognitive functioning, is similar to what others have often called the cognitive or advanced ADL domain (29)(30)(31). In subsequent field testing, we will evaluate the importance of including both the frequency and limitation dimensions to evaluate change in disability status. In some individuals, meaningful change in disability status may not be detected if only one aspect is addressed. For example, a person may not perceive a change in limitation in doing a task, but performs a more difficult task less often (32).
The disability structure that emerged from this study is similar to findings reported by Avlund and her colleagues, who have developed a measure of functional ability for healthy older persons. They used two evaluative criteria, labeled dependency and tiredness. Although the evaluative criteria were different than in the current study, domains of mobility function and personal ADLs were identified with the dependency criteria (33). Our model, however, is more complex, as personal and social factors are associated with frequency of performance, and instrumental and management factors are associated with perceived limitations in capability. The three factors (mobility and lower and upper limb functioning) that emerged from the tiredness evaluative criteria were more consistent with the structure we found for physical function (12).
The factor analytic results indicate that the two-factor solutions were more acceptable than the one-factor solutions. The results of the Rasch scaling analyses supported both the one- and two-factor solutions as reasonable hierarchical scales, although fit problems are more evident in the one-factor than the two-factor solutions. For practical measurement purposes, however, these minor fit problems were tolerable, and the hierarchies of both the one- and two-factor models provide a good representation of the continuum of disability defined by frequency of performance and limitation capability. Use of either (or both) of the total dimension and domain scores will be driven by the specific instrument application. Total dimension scores provide more precision and improved stability, whereas domain scores may be more responsive to a particular area of disability assessment. Results of the known-groups analyses confirm the item content spread of each dimension of disability and suggest that the dimensions will be able to discriminate among older persons with different levels of disability. The evidence for known-groups validity is an initial step in determining the potential of the disability component to be responsive to within-person change.
Consistent with other self-reported disability and health status measures (28)(34), test-retest correlations of the total dimension scores were in moderate to high range. In this small sample, the stability of frequency of participation was slightly lower than self-reported limitation. These data propose that a person's self-assessment of capability to perform life tasks is potentially more stable than one's perception of frequency of performance.
In summary, the disability component of the Late-Life FDI has potential to assess change in disability across a wide variety of life tasks with relatively few items. While the structure of the disability construct is complex, preliminary field-testing suggests that the disability component can discriminate among known groups that reflect the spectrum of life tasks among community-dwelling older adults. In subsequent field tests, we will evaluate the ability of the disability component of the Late-Life FDI to detect meaningful change in disability status across dimensions and role domains in a larger sample of community-dwelling older persons and compare its operating characteristics to those of other instruments commonly used in gerontologic research.
|
Acknowledgments |
|---|
Received September 11, 2001
Accepted October 25, 2001
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. J. Lenze, B. L. Rollman, M. K. Shear, M. A. Dew, B. G. Pollock, C. Ciliberti, M. Costantino, S. Snyder, P. Shi, E. Spitznagel, et al. Escitalopram for Older Adults With Generalized Anxiety Disorder: A Randomized Controlled Trial JAMA, January 21, 2009; 301(3): 295 - 303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Jette, S. M. Haley, P. Ni, S. Olarsch, and R. Moed Creating a Computer Adaptive Test Version of the Late-Life Function and Disability Instrument J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2008; 63(11): 1246 - 1256. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. L Puthoff and D. H Nielsen Relationships Among Impairments in Lower-Extremity Strength and Power, Functional Limitations, and Disability in Older Adults Physical Therapy, October 1, 2007; 87(10): 1334 - 1347. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. K LeBrasseur, S. P Sayers, M. M Ouellette, and R. A Fielding Muscle Impairments and Behavioral Factors Mediate Functional Limitations and Disability Following Stroke Physical Therapy, October 1, 2006; 86(10): 1342 - 1350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. McAuley, J. F. Konopack, K. S. Morris, R. W. Motl, L. Hu, S. E. Doerksen, and K. Rosengren Physical activity and functional limitations in older women: influence of self-efficacy. J. Gerontol. B. Psychol. Sci. Soc. Sci., September 1, 2006; 61(5): P270 - P277. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M Jette Toward a Common Language for Function, Disability, and Health Physical Therapy, May 1, 2006; 86(5): 726 - 734. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. McAuley, J. F. Konopack, R. W. Motl, K. Rosengren, and K. S. Morris Measuring Disability and Function in Older Women: Psychometric Properties of the Late-Life Function and Disability Instrument J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2005; 60(7): 901 - 909. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V Rabenda, N Burlet, O Ethgen, F Raeman, J Belaiche, and J-Y Reginster A naturalistic study of the determinants of health related quality of life improvement in osteoarthritic patients treated with non-specific non-steroidal anti-inflammatory drugs Ann Rheum Dis, May 1, 2005; 64(5): 688 - 693. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Glass, L. F. Berkman, E. F. Hiltunen, K. Furie, M. M. Glymour, M. E. Fay, and J. Ware The Families In Recovery From Stroke Trial (FIRST): Primary Study Results Psychosom Med, November 1, 2004; 66(6): 889 - 897. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Ouellette, N. K. LeBrasseur, J. F. Bean, E. Phillips, J. Stein, W. R. Frontera, and R. A. Fielding High-Intensity Resistance Training Improves Muscle Strength, Self-Reported Function, and Disability in Long-Term Stroke Survivors Stroke, June 1, 2004; 35(6): 1404 - 1409. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Pesudovs, E. Garamendi, J. P. Keeves, and D. B. Elliott The Activities of Daily Vision Scale for Cataract Surgery Outcomes: Re-evaluating Validity with Rasch Analysis Invest. Ophthalmol. Vis. Sci., July 1, 2003; 44(7): 2892 - 2899. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Farkas, A. M. Jette, S. Tennstedt, S. M. Haley, and V. Quinn Knowledge Dissemination and Utilization in Gerontology: An Organizing Framework Gerontologist, March 1, 2003; 43(90001): 47 - 56. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. J. Evans Guest Editorial Exercise as the Standard of Care for Elderly People J. Gerontol. A Biol. Sci. Med. Sci., May 1, 2002; 57(5): M260 - 261. [Full Text]
|
|
|
|
|
|
S. M. Haley, A. M. Jette, W. J. Coster, J. T. Kooyoomjian, S. Levenson, T. Heeren, and J. Ashba Late Life Function and Disability Instrument: II. Development and Evaluation of the Function Component J. Gerontol. A Biol. Sci. Med. Sci., April 1, 2002; 57(4): M217 - 222. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
2.0. See text for details.