Identifying early decline of physical function in community-dwelling older women: performance-based and self-report measures. (Research Report).

Older adults (ie, 65 years of age or older) beginning to experience a decline in physical function (the ability to perform mobility tasks, activities of daily living [ADL], and instrumental activities of daily living [IADL] that are important for achieving and maintaining an independent living status (1,2)) may fail to report this decline. Fried et al (3) referred to a progressive but unrecognized decline in physical function that precedes and often predicts onset of clinically detectable physical function decline as "preclinical disability." Preclinical disability may be represented by increased time to complete a task, modification of a task, or decreasing the frequency in which a task is performed. (3,4) Because the individual is still able to complete the task, he or she may not recognize this as a decline in function. The decline may need to reach a certain magnitude that interferes with daily functioning before the older adult recognizes it as a problem. (3,4) For this reason, researchers (3,4) have suggested that the traditional self-report measures assessing the difficulty a person has completing a task may fail to capture this "preclinical disability." However, the evidence to support the argument that self-report measures fail to match physical performance is lacking. According to Fried et al, (5) people with "preclinical disability" are at high risk for progressing to more severe disability; thus, the ability to identify this early decline in physical function is important if we plan to intervene to possibly stop the decline.

There are self-report and performance-based measures designed to measure the construct of physical function. Several researchers (2,6-12) have examined the association between performance-based and self-report measures of physical function in different populations. The results have been similar, with performance-based measures being at most moderately related to self-report measures. The lack of a stronger association between self-report and performance-based measures of physical function may be explained by the fact that they measure different aspects of the same construct (physical function). (9) With self-report measures, the individual is reporting his or her perception of ability to complete a task, whereas performance-based measures examine the person's ability to complete a task by observing his or her performance. (9) Given the idea that a person may not recognize a mild decline in physical function as a problem, (3,4) a performance-based measure of physical function may identify deficits in physical function before they are identified by a self-report measure.

Few researchers, however, have examined whether performance-based measures of physical function are more likely than self-report measures to identify deficits in physical function. Rozzini et al (8) found that a performance-based measure, the Physical Performance Test (PPT), could be used to detect deficits in physical function before they became measurable by self-reported ADL and IADL scales. In this study, Rozzini et al (8) used self-reported measures of physical function to assess independence versus dependence in completing a task and not difficulty of performing the task. Self-report measures that assess the degree of difficulty completing a task and not just the ability to complete the task (ie, independent versus dependent), in our view, may identify an early decline in physical function just as well as performance-based measures. The purpose of our study was to compare the descriptions of physical function in community-dwelling older women obtained using self-report and performance-based measures. To further explore the association between self-report and performance-based measures of physical function, we described characteristics of community-dwelling women who performed and reported similar levels of physical function and of those whose performance was different from their report.

Methods

Subjects

Subjects for our study were 170 community-dwelling older women who had participated in a long-term follow-up study of a clinical trial of walking. (13-16) The women had a mean age of 74.3 years (SD=4.3, range=56.6-83.6), a mean height of 159.5 cm (SD=6.3, range=140.1-175.3), and a mean weight of 69.3 kg (SD=15.9, range=39.5-148.0). In 1982, 229 women who were post-menopausal volunteered to participate in a 3-year randomized controlled walking intervention trial. Fourteen years after the end of the intervention, the women were invited to participate in a 1-day clinic visit that was designed to assess the long-term effects of physical activity on health status. At the 1999 follow-up, 10 of the 229 women were determined to be lost to follow-up, and 20 women were determined to be deceased. Of the remaining 199 women, 171 women participated in the clinic visit (12 women were too sick to participate, 13 women were unable to attend the clinic visit but provided information in a telephone interview, and 3 women refused to participate). Of the 171 women who attended the clinic visit, 1 woman who resided in a personal care home was excluded from the current study because she was not community dwelling.

Measures

Physical function was assessed using both performance-based and self-report measures. Two performance-based measures--gait speed and the 7-item PPT (17)--were administered by the same physical therapist. Measures of health (health status interview) and physical activity (pedometer) were used to describe the study sample. The self-report measure (the ADL, IADL, and social activity sections of the FSQ (18)) and a health status interview were administered by a research clinician and not by the physical therapist. The performance-based measures of physical function, the self-report measure of physical function, and the health status interview were administered during a single clinic visit. The order of test administration was not predetermined, but the order of testing was based on equipment and tester availability.

Gait speed. Gait speed was measured while the participants walked at a self-selected speed on an instrumented walkway, the GaitMat II * analysis system. (19) Participants completed 2 practice walks the length of the walkway (4 m) at a self-selected speed, followed by completing the walk for gait speed data collection. Participants were instructed to walk at their usual walking speed. The GaitMat II calculates gait speed by dividing the distance traversed by the time between the first and last switch closure (excluding the 1-m inactive regions at the beginning and end of the walkway).

Physical Performance Test. The PPT is a performance-based measure of physical performance of daily activities developed and tested in a sample of community-dwelling older adults. (17) The 7-item PPT consists of the following tasks: writing a sentence, simulated eating, lifting a book to a shelf above shoulder level, putting on a jacket, picking up a penny from the floor, turning 360 degrees, and walking 15.2 m (50 ft). The PPT was administered and scored following the protocol described by Reuben and Siu. (17) Six of the items (writing a sentence, simulated eating, lifting a book to a shelf, putting on a jacket, picking up a penny, and walking) are timed. Using the scoring criteria developed by Reuben and Siu, (17) the time taken to complete the item was rounded to the nearest 0.5 second and was converted to a score from 1 to 4, with a score of 4 representing the fastest (best) performance. In the original testing of the PPT, a score of 4 represented the individuals in the fastest 20% of time to complete the item, and a score of 1 represented the individuals in the slowest 20% of time for each item. If a subject was unable to complete the item, he or she was given a score of 0. For example, the item "put on and remove a jacket" was scored based on the time to completely put on and remove a jacket (unable=0, >20 s=1, 15.5-20 s=2, 10.5-15 s=3, [less than or equal to] 10 s=4). Performance of the item "turning 360 degrees" was scored by rating the continuity of the person's steps while turning (0=discontinuous steps, 2=continuous steps) and the steadiness of the person (0=unsteady, 2=steady). The scores from the individual items were summed, with the total score on the 7-item PPT ranging from 0 to 28 and with lower scores indicating poorer performance. Interrater reliability, (17) validity by comparison with accepted functional status assessments, and predictive validity for institutionalization or mortality (20) have been demonstrated in a sample of older adults of a senior housing unit, ambulatory geriatric practices, and a board-and-care facility.

Functional Status Questionnaire. The FSQ is a comprehensive self-report measure of functional status of patients receiving ambulatory care. (18) Three subscales of the FSQ relating to physical function were selected for our study: basic ADL, IADL, and social activity. The ADL subscale of the FSQ consists of questions pertaining to the amount of difficulty performing activities such as dressing, bathing, transfers, and mobility. The IADL subscale of the FSQ consists of questions pertaining to the amount of difficulty a person has shopping, using public transportation, and maintaining a household. The social activity subscale focuses more on difficulty with social interactions such as the person's ability to visit with family and friends. The subscales of the FSQ were scored according to the directions provided by Jette et al. (18) Scores on the FSQ range from 0 to 100, with higher scores representing better functional status. In a sample of older adults, the FSQ has also been shown to exhibit construct and convergent validity by comparison with health status measures such as reported bed disability days (number of days in bed all or most of the day in the past month due to illness or injury) and restricted activity days (number of days activities were limited for half of a day or more in the past month due to illness or injury). (9)

Health status measure. A structured interview was used to assess each participant's age and health status. The women were asked to report their level of satisfaction (very satisfied, satisfied, neither satisfied or dissatisfied, dissatisfied, or very dissatisfied) with their current health status and whether they currently used an assistive device for walking. In addition, the women were asked whether they were ever told by a physician that they had a list of medical conditions such as cardiovascular disease, osteoarthritis, lung disease, cancer, stroke, and other neurologic conditions. Body weight and height were measured with the participant's shoes removed. Using the height and weight measurements, body mass index (BMI) was calculated as weight (in meters) divided by height (in square kilograms).

Physical activity. A pedometer (Digi-walker SW-500 [dagger]) was used as the performance-based measure of physical activity. (21) The participants were asked to wear the pedometer for 7 consecutive days and to record in a diary the number of steps taken on each day. At the end of the 7-day period, the participant returned the diary to the study investigators. The number of steps taken during the 7-day period was averaged to achieve a single measure of physical activity (ie, mean number of steps per day). The pedometer is reported to yield valid and reliable measurements of the number of steps taken or the distance covered. (21)

Data Analysis

To compare the self-report and performance-based measures of physical function, descriptive statistics, including means, ranges, and ceiling effects, were calculated. Two questions from the self-report measure were compared with similar tasks from the performance-based measure: (1) reported difficulty walking across a room on the FSQ was compared with performance of the 15.2-m (50-ft) walk on the PPT and (2) reported difficulty with eating, dressing, or bathing on the FSQ was compared with simulated eating and putting on and taking off a jacket from the PPT. Based on their responses to the question "During the past month, how much physical difficulty did you have walking indoors, such as around your home?" from the self-report measure (FSQ), the women were classified as reporting having difficulty walking (women who reported some difficulty walking, much difficulty walking, or usually do not do because of health reasons) or as reporting having no difficulty walking. The women were also classified using the 15.2-m walk from the performance-based measure (PPT) as having difficulty walking (score of [less than or equal to] 3=taking [greater than or equal to] 15.5 seconds to walk 7.6 m [25 ft], turn, and return to the starting point) or not having difficulty walking (score of 4=taking <15.5 seconds to walk 7.6 m, turn, and return to the starting point).

For the second item, the "During the past month, how much physical difficulty did you have taking care of yourself, that is, eating, dressing, or bathing?" from the FSQ was used to classify the women as reporting having difficulty eating, dressing, or bathing (women who reported some difficulty, much difficulty, or usually do not do because of health reasons) or as reporting no difficulty eating, dressing, or bathing. The women were also classified as having difficulty eating and dressing using the "simulated eating" and "put on and take off a jacket" tasks from the PPT. Women who scored less than a 4 on either item were classified as having difficulty eating, dressing, or bathing, and women who scored 4 on both items were classified as not having difficulty eating, dressing, or bathing. Sensitivity, specificity, false positives, and false negatives of the self-report measure items for identifying difficulty with physical function as defined by the performance-based measure items were calculated. (22)

To further explore the association between self-report and performance-based measures of physical function, we describe characteristics of community-dwelling women who performed and reported similar levels of physical function (ie, walking) and of those whose performance was different from their report. The women's report of their ability to walk and their walking performance were compared using the walking question from the FSQ described above and their performance on the 15.2-m walk from the PPT. The women were classified as (1) reported having difficulty walking and demonstrated difficulty walking, (2) reported having difficulty walking and did not demonstrate difficulty walking, (3) reported having no difficulty walking and demonstrated difficulty walking, and (4) reported having no difficulty walking and did not demonstrate difficulty walking.

Results

Eighty-three percent of the women reported that they were satisfied with their health, and 21% of the women reported having 2 or more chronic conditions such as arthritis, cardiovascular disease, diabetes, or lung disease. A large percentage (>60%) of the women scored at the ceiling for the self-report measures of function (Tab. 1). For the performance-based measures of function, only 7% of the women achieved a perfect score on the PPT (ceiling effect), and 30% of the women scored at the ceiling (>1.2 m/s) for gait speed. Because gait speed is a continuous measure, technically there would not necessarily be a ceiling value. To examine the ceiling effect of gait speed, we used walking faster than 1.2 m/s, which is the desired gait speed of community-dwelling adults. (23)

The subjects' responses to questions from the self-report measures of walking and feeding and dressing demonstrated high specificity (0.97-0.98) but low sensitivity (0.08-0.09) when compared with the performance-based identified difficulties (Tab. 2). The high specificity means that people who did not have problems based on the PPT were identified as not having problems (low false positive rate) by self-report. (22) For example, 37 women demonstrated no difficulty eating and dressing, and 36 of the women also reported having no difficulty with similar tasks. The low sensitivity means that the self-report measure has a "good" chance of identifying someone as not having a problem when that person actually does have difficulty performing the tasks (high false negative rate). (22) For example, of the 133 women who demonstrated difficulty eating or dressing, only 10 women reported having difficulty with similar tasks.

More than half of the women (n=99) performed and reported similar levels of walking ability (Tab. 3). Of the women whose performance was different from their report, 69 failed to report having difficulty walking when they actually had difficulty walking. Only 2 women reported having difficulty walking when they did not have difficulty walking. Because the groups had unequal numbers of women, we report only descriptive statistics. The women who performed and reported similar levels of physical function (n=99) had on average a lower BMI than the women whose performance was different from their report (26.1 and 26.1 versus 28.5 and 33.0). The 2 women who reported having difficulty walking when they had no difficulty walking tended to be overweight women (BMI: 33.0 versus 26.1, 28.5, and 26.1) who were likely to have chronic conditions (50% versus 14%, 18%, and 25%), but were satisfied with their health (Tab. 3).

When comparing the women who did not report having difficulty walking by their ability to walk, there are several notable differences (Tab. 3). Compared with women who did not report having difficulty walking and did not have difficulty walking, women who did not report having difficulty walking but had difficulty walking were slightly older (75.3 years versus 73.5 years), had a higher BMI (28.5 versus 26.1), and were more likely to use an assistive device (12% versus 2%). Though the 2 groups of women reported similar levels of function on the FSQ, the women who did not report having difficulty walking but had difficulty walking performed poorer on the performance-based measures of gait speed (0.97 versus 1.24 m/s) and the PPT (23.0 versus 25.8).

Physical activity, as measured by the pedometer, appeared to be related to a person's perception (self-report measure) and ability (performance-based measure) to walk (Tab. 3). Women who perceived that they had difficulty walking were less physically active than the women who perceived that they did not have difficulty walking (1,806 and 2,906 steps/d versus 3,755 and 5,903 steps/d). Women who had difficulty walking but perceived that they did not have difficulty walking (n=69) were more physically active than women who did not have difficulty walking but perceived that they did have difficulty walking (n=2) (3,755 and 2906 steps/d, respectively). Similar results were obtained when the women were described based on their response on the self-report measure and their performance on the performance-based measure of ADL (data not shown).

Discussion and Conclusion

In a sample of community-dwelling older women, a commonly used self-report measure of functional status (FSQ) had a moderate to severe ceiling effect, whereas the 2 performance-based measures (PPT and gait speed) were much less likely to have ceiling effects. With a high percentage of participants having the maximum score on the FSQ, indicating no difficulty, and a much lower percentage of participants having the maximum score on the performance-based measures, we suggest that the performance-based measures are more likely than the self-report measure to identify deficits in physical function. In our sample of relatively high-functioning, community-dwelling older women, the FSQ had a greater ceiling effect (ADL=77%, IADL=61%) than has previously been reported. (9,10) In other samples of community-dwelling older adults of similar mean age (76 years), less severe ceiling effects of 41% and 64% for the ADL section of the FSQ and 16% and 23% for the IADL section of the FSQ have been reported. (7,9) Thus, in a relatively high-functioning cohort of older people, the FSQ may fail to provide adequate information regarding mild deficits in physical function.

In a high-functioning sample, sensitive measures are needed to identify early declines in physical function (ie, what has been called "preclinical disability"). From our data, it appears that women who have minor problems with functional tasks (slowed performance) are not likely to report these problems as having difficulty performing the task. Our results build on the earlier findings by Rozzini and colleagues (8) that suggest that performance-based measures of function may detect deficits in physical function before they become measurable by self-reported ADL and IADL scales. We were able to show that performance-based measures were more likely to identify a deficit in physical function than a self-reported measure, which assessed the degree of difficulty a person has completing a task.

We are not suggesting that one type of measure, self-report or performance-based, is superior to the other. We argue that when anyone attempts to identify early, mild deficits in physical function, the format of the self-report measure should be taken into consideration. In order to measure any early decline in function with a self-report measure of function, it may be important to inquire about whether a person has modified his or her physical performance or whether there has been a change in the frequency of performing the task in addition to the amount of difficulty performing the task. (4) In a cross-sectional study, Fried et al (4) demonstrated the ability to identify people with early decline in physical function using a self-report measure that assessed task modification and change in frequency of task performance (both characteristics of "preclinical disability"). Recently, this self-report method of identifying disability was found to be predictive of incident (new onset of) mobility difficulty 18 months later in older women. (5) Interestingly, the self-report method of identifying disability gave results just as well if not better than the performance-based measure of gait speed in predicting future mobility difficulty in older adults who were functioning at a high level (they required no modification in task performance and had no difficulty performing tasks). (5)

We chose to define a deficit in physical performance as scoring below 4 (highest score) on the PPT. Scores on the PPT are based on timing of the task, and an individual who has a score of 3 instead of 4 requires a greater amount of time to complete the task. (17) By using increased time to complete a task, we used a measure that complies with what Fried and colleagues (3,4) called "preclinical disability" (ie, increased time to complete a task, modification of a task, or decreasing the frequency in which a task is performed). In addition, lower scores on the PPT, representing increased time to complete the tasks, are predictive of institutionalization and death in older people. (20) Therefore, if we are attempting to identify early indicators of disability, a low score on the PPT (or increased time needed to complete a task), in our view, would be a good indicator.

Some people may argue that because a person was able to complete a task, but just took increased time to do so, this should not be considered a deficit in physical function. However, we feel that if the goal is to reduce or delay disability, then identifying early decline, not just inabilities (deficits), in physical function is important. We contend that, at this early stage of decline, interventions to improve physical function may be less costly, less intensive, and more effective. For example, people with this minimal deficit may need to modify their lifestyle through the use of health promotion and education programs, such as establishing regular physical activity (walking) programs, to reduce the risk for future disability. (24,25) In contrast, people who demonstrate a moderate or severe decline of physical function may require extensive rehabilitation to improve physical function first, before continuing with a physical activity program. Rehabilitation of people with deficits in physical function prior to initiation of a regular physical activity (walking) program is more costly than initiating a regular physical activity (walking) program alone, and the success of the rehabilitation on physical function and future disability has yet to be demonstrated. (26,27)

Because we did not consider one type of measure, self-report or performance-based, to be superior to the other in our study, we compared the sensitivity and specificity of the 2 types of measures. Other authors (9) have suggested that because self-report and performance-based scales measure different aspects of the construct of physical function (perception versus ability), each measure provides unique and useful information in the assessment of older people. Given what we know about older people and the recognition of disability, it is not surprising that several older people did not report difficulty when, based on our definition (ie, scoring <4 on the 15.2-m walk item of the PPT), they had difficulty walking. Just because these people walked slowly, does not mean that they will perceive this as having difficulty walking.

Even when using this strict definition of not having difficulty walking (ie, scoring 4 on the 15.2-m walk item of the PPT) when comparing the self-report and performance-based measures, more than half of the women in our study (n=99) performed and reported similar levels of physical function. Of the women whose performance was different from their report, they were more likely to over-report their ability (fail to report having difficulty walking when they actually had difficulty walking, n=69) than to under-report their ability (report having difficulty walking when they did not have difficulty walking, n=2). The women who under-reported their ability (reported having difficulty walking when they did not have difficulty walking) may have been reflecting the effort involved in completing the task. Being overweight (mean BMI=33.0), using an assistive device, and having multiple chronic conditions, the women are likely to use greater relative energy expenditure to complete the task. That is, they may complete the task in a "normal" amount of time, but they are more likely to experience shortness of breath or increased heart rate compared with people without known pathology or impairments. However, given the small numbers (only 2 women under-reported their ability), any statements made about these women must be interpreted cautiously. In addition, the women who over-reported their ability (reported having no difficulty walking but had difficulty walking) were less active then the women who consistently reported their performance (reported having no difficulty walking and did not have difficulty walking). Walking less often, these women may not have an accurate perception of their ability.

Our finding that a person's perception of his or her ability to walk appears to be related to physical activity, with women who perceive that they have difficulty walking being less physically active than women who do not perceive that they have difficulty walking, is similar to the findings of other researchers. (28) Identification of factors associated with the perception of walking difficulty is important if interventions to increase physical activity through a walking program are planned.

In our study of community-dwelling older women, the performance-based measures of the PPT and gait speed identified deficits in physical function not indicated by the FSQ, a self-report measure of function. Among older adults functioning at a high level of physical performance, early identification of minor problems in physical functioning using performance-based measures may provide the opportunity for early intervention to reduce physical disability.

Table 1.
Description of Self-Report and Performance-Based Measures (N= 170) (a)

                                     [bar]X   SD     Median

Self-report measure (FSQ)
  ADL (0-100)                        96.2      7.9   100
  IADL (0-100)                       93.6     10.4   100
  Social activity (0-100)            98.5      6.5   100

Performance-based measures
  Physical Performance Test (0-28)   24.4      2.6    25
  Gait speed (m/s)                    1.12    0.21     1.13

                                                 %
                                     Range       Ceiling

Self-report measure (FSQ)
  ADL (0-100)                        66.7-100    77
  IADL (0-100)                       66.7-100    61
  Social activity (0-100)            55.6-100    94

Performance-based measures
  Physical Performance Test (0-28)      9-28      7
  Gait speed (m/s)                   0.33-1.72   30 (b)

(a) FSQ=Functional Status Questionnaire, ADL=activities of daily
living, IADL=instrumental activities of daily living.

(b) Ceiling effect for gait speed defined as >1.2 m/s based on desired
gait speed of older adults. (23)
Table 2.
Comparison of Self-Report (Functional Status Questionnaire) and
Performance-Based (Physical Performance Test) Measures

                    Performance
Walking             Based
Self-Report         <4       4    Sensitivity   Specificity

Some difficulty      7       2    0.09          0.98
No difficulty       69      92
Total               76      94

Eating, Dressing,   Performance
and Bathing         Based (a)
Self-Report         <8     8      Sensitivity   Specificity

Some difficulty      10    1      0.08          0.97
No difficulty       123   36
Total               133   37

Walking              False      False
Self-Report          Positive   Negative

Some difficulty      0.02       0.91
No difficulty
Total

Eating, Dressing,
and Bathing          False      False
Self-Report          Positive   Negative

Some difficulty      0.03       0.92
No difficulty
Total

(a) Combined simulated eating and dressing items from Physical
Performance Test.
Table 3.
Description of Women by Self-Report and Performance-Based Status for
Walking (a)

Reported difficulty
  walking (b)          Yes          Yes          No          No
Had difficulty
  walking (c)          Yes          No           Yes         No
                       n=7          n=2          n=69        n=92

Age (y), [bar]X        74.8 [+ or   73.6 [+ or   75.3 [+ or  73.5 [+ or
  [+ or -] SD            -] 5.7       -] 5.8       -] 4.5      -] 3.9
Body mass index        26.1 [+ or   33.0 [+ or   28.5 [+ or  26.1 [+ or
                         -] 4.2       -] 3.4       -] 5.4      -] 5.5
Satisfied with         43%          100%         81%         87%
  health
Reported               14%           50%          25%        18%
  [less than or
  equal to] 2
  chronic
  conditions
Used an assistive      57%          50%          12%         2%
   device
Median no. of steps
  taken per day
  (25th and 75th
  percentiles)         1,806        2,906        3,755       5,903
                       (1,038,      (2,243,      (2,449,     (4,594,
                         5,148)       3,677)       5,822)      7,786)
Gait speed (m/s),      0.85 [+ or   1.07 [+ or   0.97 [+ or  1.24 [+ or
  [bar]X [+ or -] SD     -] 0.28      -] 0.04      -] 0.15     -] 0.15
7-item PPT,            20.3 [+ or   26.5 [+ or   23.0 [+ or  25.8 [+ or
  [bar]X [+ or -] SD     -] 5.4       -]  0.7      - ] 2.1     -] 1.6
FSQ-ADL,               69.8 [+ or   72.2 [+ or   97.1 [+ or  98.2 [+ or
  [bar]X [+ or -] SD     -] 5.4       -] 7.9       -] 5.9      -] 4.1
FSQ-IADL,              73.2 [+ or   76.7 [+ or   91.9 [+ or  96.8 [+ or
  [bar]X [+ or -] SD     -] 9.3       -] 4.7       -] 11.3     -] 7.0
FSQ-social activity,   86.5 [+ or   100 [+ or    97.8 [+ or  99.9 [+ or
  [bar]X [+ or -] SD     -] 15.0      -] 0         -]  8.1     -] 1.2

(a) PPT=Physical Performance Test, FSQ=Functional Status Questionnaire,
ADL=activities of daily living, IADL=instrumental activities of daily
living.

(b) Based on question "During the past month, how much physical
difficulty did you have walking indoors, such as around your home?"
from the ADL subscale of the FSQ. Women who reported some difficulty,
much difficulty, or usually do not do because of health reasons were
classified as reported having difficulty walking.

(c) Based on the 15.2-m (50-ft) walk from the PPT, women with scores of
[greater than or equal to] 3 (ie, taking [less than or equal to] 15.5 s
to walk 7.6 m [25 ft], turn, and return to the starting point) were
classified as having difficulty walking.

This study was approved by the Biomedical Institutional Review Board of the University of Pittsburgh.

This study was funded by the National Institute on Aging (Grant AG 14753). Dr Brach was supported, in part, by the Foundation for Physical Therapy, the Section on Geriatrics of the American Physical Therapy Association, and the National Institutes of Health (Public Health Service Grant TG32AG00181).

This study was presented, in part, at the Combined Sections Meeting of the American Physical Therapy Association; February 14-18, 2001; San Antonio, Tex.

This article was submitted April 2, 2001, and was accepted September 10, 2001.

* EQ Inc, PO Box 16, Chalfont, PA 18914.

[dagger] New Lifestyles Inc, 5900 Larson Ave, Kansas City, MO 64133.

References

(1) Guralnik JM, Simonsick EM. Physical disability in older Americans. J Gerontol. 1993;48:3-10.

(2) Cress ME, Schechtman KB, Mulrow CD, et al. Relationship between physical performance and self-perceived physical function. J Am Geriatr Soc. 1995;43:93-101.

(3) Fried LP, Herdman SJ, Kuhn KE, et al. Preclinical disability: hypotheses about the bottom of the iceberg. J Aging Health. 1991;3:285-300.

(4) Fried LP, Bandeen-Roche K, Williamson JD, et al. Functional decline in older adults: expanding methods of ascertainment. J Gerontol A Biol Sci Med Sci. 1996;51:M206-M214.

(5) Fried LP, Bandeen-Roche K, Chaves PHM, Johnson BA. Preclinical mobility disability predicts incident mobility disability in older women. J Gerotonl A Biol Sci Med Sci. 2000;55:M43-M52.

(6) Sager MA, Dunham NC, Schwantes A, et al. Measurement of activities of daily living in hospitalized elderly: a comparison of self-report and performance-based methods. J Am Geriatr Soc. 1992;40:457-462.

(7) Elam JT, Graney MJ, Beaver T, et al. Comparison of subjective ratings of function with observed functional ability of frail older persons. Am J Public Health. 1991;81:1127-1130.

(8) Rozzini R, Frisoni GB, Ferrucci L, et al. The effect of chronic diseases on physical function: comparison between activities of daily living scales and the Physical Performance Test. Age Ageing. 1997;26:281-287.

(9) Reuben DB, Valle LA, Hays RD, Siu AL. Measuring physical function in community-dwelling older persons: a comparison of self-administered, interviewer-administered, and performance-based measures. J Am Geriatr Soc. 1995;43:17-23.

(10) Sherman SE, Reuben DB. Measures of functional status in community-dwelling elders. J Gen Intern Med. 1998;13:817-823.

(11) Ferrer M, Lamarca R, Orfila F, Alonso J. Comparison of performance-based and self-rated functional capacity in Spanish elderly. Am J Epidemiol. 1999;149:228-235.

(12) Merrill SS, Seeman TE, Kasl SV, Berkman LF. Gender differences in the comparison of self-reported disability and performance measures. J Gerontol A Biol Sci Med Sci. 1997;52:M19-M26.

(13) Cauley JA, Kriska AM, LaPorte RE, et al. A two-year randomized exercise trial in older women: effects on HDL-cholesterol. Atherosclerosis. 1987;66:247-258.

(14) Kriska AM, Bayles C, Cauley JA, et al. A randomized exercise trial in older women: increased activity over two years and the factors associated with compliance. Med Sci Sports Exerc. 1986;18:557-562.

(15) Pereira MA, Kriska AM, Day RD, et al. A randomized walking trial in postmenopausal women: effects on physical activity and health 10 years later. Arch Intern Med. 1998;158:1695-1701.

(16) Sandler RB, Cauley JA, Hom DL, et al. The effects of walking on the cross-sectional dimensions of the radius in postmenopausal women. Calcif Tissue Int. 1987;41:65-69.

(17) Reuben DB, Siu AL. An objective measure of physical function of elderly outpatients: the Physical Performance Test. J Am Geriatr Soc. 1990;38:1105-1112.

(18) Jette Am, Davies AR, Cleary PD, et al. The Functional Status Questionnaire: reliability and validity when used in primary care. J Gen Intern Med. 1986;1:143-149.

(19) Walsh JP. Foot fall measurement technology. In: Craik RL, Oatis CA, eds. Gait Analysis: Theory and Application. St Louis, Mo: CV Mosby Co; 1995:125-142.

(20) Reuben DB, Siu AL, Kimpau S. The predictive validity of self-report and performance-based measures of function and health. J Gerontol. 1992;47:M106-M110.

(21) Bassett DR Jr, Ainsworth BE, Leggett SR, et al. Accuracy of five electronic pedometers for measuring distance walked. Med Sci Sports Exerc. 1996;28:1071-1077.

(22) Sackett DL, Straus SE, Richardson WS, et al. Evidence-Based Medicine: How to Practice and Teach EBM. 2nd ed. London, England: Churchill Livingstone; 2000:67-93.

(23) Langlois JA, Keyl PM, Guralnik JM, et al. Characteristics of older pedestrians who have difficulty crossing the street. Am J Public Health. 1997;87:393-397.

(24) Cress ME, Buchner DM, Questad KA, et al. Exercise: effects on physical functional performance in independent older adults. J Gerontol A Biol Sci Med Sci. 1999;54:M242-M248.

(25) Rooks DS, Kiel DP, Parsons C, Hayes WC. Self-paced resistance training and walking exercise in community-dwelling older adults: effects on neuromotor performance. J Gerontol A Biol Sci Med Sci. 1997;52:M161-M168.

(26) King AC, Rejeski WJ, Buchner DM. Physical activity interventions targeting older adults: a critical review and recommendations. Am J Prev Med. 1998;15:316-333.

(27) Keysor JJ, Jette AM. Have we oversold the benefit of late-life exercise? J Gerontol A Biol Sci Med Sci. 2001;56:M412-M423.

(28) Simonsick EM, Guralnik JM, Fried LP. Who walks? Factors associated with walking behavior in disabled older women with and without self-reported walking difficulty. J Am Geriatr Soc. 1999;47:672-680.

JS Brach, PT, PhD, GCS, is Assistant Professor, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, 6035 Forbes Tower, Pittsburgh, PA 15260 (USA) ([e-mail]). Address all correspondence to Dr Brach.

JM VanSwearingen, PT, PhD, is Associate Professor, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh.

AB Newman, MD, MPH, is Associate Professor, Department of Geriatric Medicine, School of Medicine, and Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh.

AM Kriska, PhD, FACSM, is Associate Professor, Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh.

All authors provided concept/research design. Dr Brach and Dr VanSwearingen provided writing. Dr Brach provided data collection and analysis, and Dr Kriska provided subjects, fund procurement, and facilities/equipment. Dr VanSwearingen, Dr Newman, and Dr Kriska provided consultation (including review of manuscript before submission). The authors thank the women who participated in this study, whose dedication made this long-term research project possible.