Muscle Force and Range of Motion as Predictors of Function in Older Adults.

About 10% of the nondisabled adult population aged 75 years and older lose independence in one or more basic activities of daily living (ADL) each year.[1] According to government data, an estimated 20% of the population over the age of 65 years has difficulty in one or more ADL tasks.[2] Limitations in basic self-care tasks such as bathing, dressing, eating, and toileting increase dramatically with age. Similar trends are noted with instrumental activities of daily living (IADL) such as shopping, preparing meals, and doing housework.[3]

With the aging of the American population, the prevention of disability has become a major focus in geriatrics. Increasing research supports the need to develop and use mechanisms for early identification of individuals who are at risk for functional decline and to establish interventions to reverse or slow the progression toward disability.[4-6]

A growing body of research indicates that musculoskeletal impairment and functional limitations contribute to disability.[7-11] Guralnik and colleagues[9] found that poor performance on physical tests of balance, walking speed, and chair rise time predicted onset of self-reported limitations in the ability to ambulate 0.8 km (0.5 mile). Lawrence and Jette[11] further examined the relationships between the onset of lower-extremity mobility problems such as difficulty walking 0.4 km (0.25 mile) or climbing steps and the onset of difficulty in performing IADL tasks. Jette et al[8] and Hughes and colleagues[10] found that upper-extremity joint impairments, including pain, limitations in range of motion (ROM), and joint deformity, are related to self-reported loss of independence in basic ADL tasks.

A number of researchers have examined the effects of various impairments on the performance of functional tasks. Bergstrom and colleagues[12] found that decreased lower-extremity ROM was associated with self-reported difficulty in functional mobility, such as rising from a chair, stair climbing, and the need for assistive devices during ambulation. Woolley and colleagues[13] found that knee extension force and subject rating of pain while rising from the floor accounted for 28% of the variability in timed performance of this task in subjects with osteoarthritis. Knee flexion and extension force, in combination with reported function and body weight, explained 47% of the variance in stair ascension time.[13] Other researchers found that lower-extremity force is a determinant of the minimum chair height from which an individual can rise[14] and the speed of rising from a chair.[15]

Much of the previous research on functional decline has been based on subject self-reports of disability.[4,5,7-11] Although subject self-reports are commonly used to measure and document disability, they may not be the best method of measuring the functional abilities of older adults. There is evidence that older adults often underreport their disabilities or minimize the extent of their limitations,[16-18] especially when the subject has dementia.[19] Reports by proxies, such as spouses, friends, or caregivers, also often provide inaccurate portrayals of an individual's functional ability.[18-20] When completing self-report instruments, subjects may not distinguish between activities they believe they are capable of doing and what they actually do at home, thereby making it difficult to determine their "hypothetical" abilities as compared with their actual performance.[21]

The alternative to using self-reported functional abilities is the observation and measurement of an individual's performance of functional tasks. We believe performance-based measures of function are time consuming and may not provide a completely accurate assessment of an individual's functional ability. For example, an individual may perform very well in the controlled laboratory environment but be unable to do the same or similar tasks in a home setting. However, performance-based measures are more likely to reflect an individual's actual abilities than are self-reports or proxy reports of function.[20]

Although there is some evidence that musculoskeletal impairments have an impact on function and disability in older adults, the types of impairments that most affect overall function have not been identified. This information could be helpful in designing interventions to prevent or slow the disablement process. For those individuals who have functional limitations, this information would be helpful to optimize efforts to restore function by focusing on the most functionally limiting musculoskeletal impairments. The primary purpose of this study was to identify some extremity musculoskeletal impairments that best predict functional performance in older adults. A second purpose was to further assess the validity of measurements obtained with the Physical Performance Test (PPT) for predicting disability in an older population.

Method

Subjects

Subjects were recruited from senior housing communities offering various levels of care. Eighty subjects (22 men and 58 women) gave their informed consent and participated in full data collection. Subjects were older adults with no acute medical conditions. Subject age averaged 81 years (SD=6.73, range=60-92). Thirty-eight subjects were residents of assisted living or skilled nursing facilities and were dependent in one or more ADL tasks. Forty-two subjects were residents of a continuing care retirement community dwelling in independent apartments and cottages. These subjects received some services, such as meals and light housekeeping, through the community, but they typically were independent in basic ADL tasks. One subject from this community was dependent in self-care activities and resided in the independent living apartments only due to the assistance of his spouse. This individual's data were excluded from the analysis. Subjects who resided in assisted living or skilled nursing facilities were classified as dependent, and subjects residing in independent apartments and cottages were defined as independent.

Instruments

Function was measured using the PPT.[22] The PPT is a standardized 7- or 9-item test that examines tasks related to both ADL and IADL performance and that requires timed performance of most items.[22] Subject performance on each item is categorized on a predefined 5-point scale (0-4), with the lowest scores indicating the poorest performance. For timed items, a score of 0 is assigned if the task cannot be completed or takes a very long time to complete, and a score of 4 is assigned for rapid completion of the task. Prior research with the PPT has demonstrated construct validity due to its correlation with self-reported measures of ADL and IADL performance.[22] The PPT has also been found to be predictive of the major health outcomes of death and nursing home placement.[23] Unpublished research indicates that PPT scores may be useful in differentiating between community-dwelling elderly people and those requiring a higher level of care.[24] The instrument has been shown to have acceptable reliability. In a study by Reuben and Siu,[22] the 7-item version of the PPT yielded a Cronbach alpha coefficient of .79 and a correlation coefficient of .93 for interrater reliability.

Due to safety concerns for the most frail of the subjects, the 7-item version of the PPT, which eliminates items requiring stair climbing, was used in our study. The tasks in the 7-item version of the PPT were writing a sentence, simulated eating, putting on and removing a jacket, lifting a book to place it on a shelf, picking up a penny, turning 360 degrees, and walking 15.2 m (50 ft). Testers followed the testing protocol established in the initial development of the PPT.[22] Each subject was given verbal instructions for each task and was allowed up to 2 trials to complete each of the 7 tasks. Six of the 7 items were timed from the command "go" until the task was completed, using a standard stopwatch. The untimed item, turning 360 degrees, was scored using a standardized scale that graded the continuity of steps and the steadiness of the subject during task performance. For the final 2 items, turning 360 degrees and the 15.2-m walk test, subjects were allowed to use assistive devices, if needed.

The musculoskeletal impairments measured in this study were selected based on the work by Gerety and colleagues in the development of the Physical Disability Index.[25] The variables measured are shown in the Figure. Five upper-extremity and 3 lower-extremity measurements of muscle force were taken bilaterally. Muscle force was measured with PowerTrack II Commander handheld dynamometers(*) and recorded in pounds. According to the manufacturer, the 2 dynamometers used in this study were factory calibrated prior to data collection. All force measurements were taken in the midrange of motion with the subject positioned supine, with the exception of handgrip and knee extension force, which were tested at midrange with the subject in a sitting position. Testing methods were consistent with those used previously.[26] Intratester test-retest reliability was assessed by remeasuring 10 subjects' muscle force within 1 week of the original measurements. Pearson product moment correlation coefficients ranged from .86 to .99. For each muscle group, the measurement was taken 3 times, and the average measurement of force was used for data analysis. In an effort to prevent fatigue, a 1-minute rest period was allowed between force measurements.

Figure. Extremity muscle force and range of motion variables included in analysis.

Upper-Extremity Range                  Lower-Extremity Range
of Motion                              of Motion

Shoulder flexion                       Hip flexion
Shoulder medial (internal) rotation    Knee flexion
Shoulder lateral (external) rotation   Knee extension
Elbow flexion                          Ankle dorsiflexion
Elbow extension                        Ankle plantar flexion
Supination
Pronation

Upper-Extremity Muscle Force           Lower-Extremity Muscle Force

Shoulder flexion                       Hip flexion
Shoulder medial rotation               Knee extension
Shoulder lateral rotation              Ankle dorsiflexion
Elbow
Handgrip

Passive range of motion measurements were taken for 7 upper-extremity and 5 lower-extremity motions. Subjects' limbs were taken passively through the motion to be measured several times before positioning the joint at the end-range of motion. Measurements were taken with a standard goniometer with subjects positioned supine. Measurements were read by a single tester at each session, with a second tester providing stabilization or positioning of the joint. Intrarater reliability was assessed by remeasuring 10 subjects' ROM within 1 week of the original measurements. Pearson product moment correlation coefficients ranged from .70 to .93.

Subjects were tested in 2 sessions of 30 to 45 minutes' duration each, with sessions scheduled 1 week apart. The PPT was administered in 1 session, and the muscle force and ROM measurements were taken in the other session. The order of testing varied. Subjects were given time to rest between task performances. Rest times were not recorded or standardized, and testing resumed when the subject indicated that he or she was ready to continue.

Data Analysis

Because muscle force, ROM, and function were measured using different scales and units of measurement, data were converted to a percentage of the "maximum" score. The PPT scores also were converted to a percentage of the "maximum" score. Subjects' ROM measurements were converted to the percentage of normal ROM, using the American Academy of Orthopaedic Surgeons definition of normal ROM for each joint.[27] Because no maximum values for muscle force were available for the conversion of muscle force measurements, the force measurements were converted to the percentage of a maximum value defined by the study population, as follows: maximum force for a given motion was taken as the average value for that motion, plus 2 standard deviations. Force measurements were standardized separately for male and female subjects, based on the methods used by Gerety and colleagues.[25] Measurements that exceeded the maximum value, as defined above, were assigned a score of 100. Using these conversion methods, theoretically scores for all measures could range from 0 to 100, with 100 representing the best possible performance.

Due to the large number of joints assessed and concerns regarding the high correlations among the force variables, aggregate scores representing the degree of impairment in a given area were created for analysis. Aggregate scores were created by averaging the standardized values for upper-extremity force, upper-extremity ROM, lower-extremity force, and lower-extremity ROM.[25] The variables that were averaged to create these aggregate scores are shown in the Figure.

All data analysis was performed using SPSS Professional Statistics 9.0([dagger]) on a desktop computer. Descriptive statistics were computed for the sample as a whole and according to level of care. To examine the predictive validity of data obtained for the PPT, logistic regression was used to determine whether PPT scores could predict the level of care received by study subjects.

Multiple regression analysis using ordinary least squares was performed to determine the contributions of impairments to function. The PPT score was regressed on subject age, sex, upper-extremity force, upper-extremity ROM, lower-extremity force, and lower-extremity ROM. A forward stepwise method was used to identify predictor variables. The criterion for inclusion in the model was set at an alpha level of .05, and the exclusion criterion was set at an alpha level of .10. Separate regressions were run for the full group and then for the subgroups according to level of care. Independent-samples t tests were used to test for differences between those subjects in independent and assisted living arrangements on each variable.

Results

Table 1 shows the means, standard deviations, and ranges for each variable for the entire sample. Table 2 contains the means and standard deviations for the sample divided into groups based on level of care. Subjects in independent living arrangements were younger, had higher scores on tests of function, and had better scores on extremity impairment measures examined in this study compared with subjects in the assisted living group. Subjects showed greater deficits and greater variability in measurements of extremity force than in ROM. The correlations among independent variables are shown in Table 3.

Table 1. Descriptive Statistics of the Sample (N=80)

Measure                  [bar]X   SD      Minimum   Maximum

Age (y)                  81        6.73   66.0      96.0

Lower-extremity
ROM(a) ([degrees])       89        6.71   73.1      99.1

Upper-extremity
ROM ([degrees])          91        6.04   74.7      100

Lower-extermity muscle
force (lb)               51       19.5    12.6      100

Upper-extremity muscle
force (lb)               54       18.0    23.3      97.0

(a) ROM=range of motion.

Table 2. Differences Between Subjects in Independent Living and Assisted Living Arrangements(a)

                      Assisted      Independent
                       Living         Living
                       (n-38)         (n=42)

Measure             [bar]X   SD     [bar]X   SD     t

Age (y)             84        6.6   78        5.6     4.5(**)

PPT                 34       23.1   90       13.9   -13.1(**)

Upper-extremity
muscle Force (lb)   43       12.9   63       16.6    -6.0(**)

Lower-extremity
muscle force (lb)   38       14.1   63       15.9    -7.4(**)

Upper-extremity
ROM ([degrees])     89        5.9   93        5.5    -3.2(*)

Lower-extremity
ROM ([degrees])     84        6.8   93        3.4    -7.4(**)

(a) PPT=Physical Performance Test score, ROM=range of motion,

asterisk (*) indicates significant at P [is less than or equal to] .01,

double asterisk (**) indicates significant at P [is less than or equal to] .001;

df=78 (independent samples t test).

Table 3. Pearson Product Moment Correlation Coefficients Among Variables

Measure(a)                     Age    PPT

Age                            1.00
PPT                            -.56   1.00
Upper-extremity muscle force   -.43    .69
Lower-extremity muscle force   -.44    .71
Upper-extremity ROM            -.33    .43
Lower-extremity ROM            -.31    .77

                                 Muscle Force

                               Upper       Lower
Measure(a)                     Extremity   Extremity

Age
PPT
Upper-extremity muscle force   1.00
Lower-extremity muscle force    .88        1.00
Upper-extremity ROM             .37         .24
Lower-extremity ROM             .55         .52

                                 Range of Motion

                               Upper       Lower
Measure(a)                     Extremity   Extremity

Age
PPT
Upper-extremity muscle force
Lower-extremity muscle force
Upper-extremity ROM            1.00
Lower-extremity ROM             .46        1.00

(a) Significant at P [is less than or equal to] .05 (independent samples t test).

The first regression with the full model indicated that the functional level of subjects was predicted best by age, lower-extremity ROM, and lower-extremity muscle force (Tab. 4). Lower-extremity ROM was the first independent variable to emerge as a predictor. The adjusted [R.sup.2] values changed considerably, from .59 to .77, from the first through the third steps.

Table 4. Results of Regression Analyses for Predicting Physical Performance Scores in Residents of Senior Housing (N=80)

                         Standardized

Measure(a)          [Beta]   SE     [Beta]   t

Step 1(b)
  Lower-extremity
  ROM                3.86    0.36    0.77    10.75

Step 2(c)
  Lower-extremity
    ROM              2.77    0.35    0.55     7.95
  Lower-extremity
    muscle force     0.72    0.12    0.42     6.01

Step 3(d)
  Lower-extremity
    ROM              2.62    0.32    0.53     8.28
  Lower-extremity
    muscle force     0.56    0.12    0.32     4.82

Age                 -1.28    0.30   -0.26    -4.26

(a) Significant at P [is less than or equal to] .001 (independent samples t test), ROM=range of motion.

(b) Adjusted [R.sup.2]=.59.

(c) Adjusted [R.sup.2]=.72, [R.sup.2] change=.60, P [is less than or equal to] .001.

(d) Adjusted [R.sup.2]=.77, [R.sup.2] change=.05, P [is less than or equal to] .001.

Because the subjects in the 2 groups had different mean scores on all variables studied, separate regression analyses were performed for these groups. These results are shown in Tables 5 and 6. In both groups, the adjusted [R.sup.2] values changed at each step of analysis. Lower-extremity ROM and upper-extremity muscle force were the best predictors for those subjects living in an assisted living environment (Tab. 5). For those subjects in an independent living arrangement, the strongest predictors of function were age and upper-extremity ROM (Tab. 6).

Table 5. Results of Regression Analyses for Predicting Physical Performance Scores in Residents of an Assisted Living Arrangement (n=38)

                         Standardized

Measure(a)          [Beta]   SE     [Beta]   t

Step 1(b)
  Lower-extremity
    ROM             2.47     0.39   0.73     6.37

Step 2(c)
  Lower-extremity
    ROM             1.70     0.36   0.50     4.71
  Upper-extremity
    muscle force    0.84     0.19   0.47     4.41

(a) Significant at P [is less than or equal to] .001 (independent samples t test), ROM=range of motion.

(b) Adjusted [R.sup.2]=.52.

(c) Adjusted [R.sup.2]=.68, [R.sup.2] change=.17, P [is less than or equal to] .001.

Table 6. Results of Regression Analyses for Predicting Physical Performance Scores in Residents of an Independent Living Arrangement (n=42)

                          Standardized

Measure(a)          [Beta]   SE     [Beta]   t

Step 1(b)
  Age(**)           -1.64    0.30   -0.66    -5.50

Step 2(c)
  Age(**)           -1.53    0.29   -0.61    -5.25
  Upper-extremity
    ROM(*)           0.63    0.29    0.25     2.17

(a) Asterisk (*) indicates P [is less than or equal to] .05, double asterisk (**) indicates significant at P [is less than or equal to] .001 (independent samples t test), ROM=range of motion.

(b) Adjusted [R.sup.2]=.42.

(c) Adjusted [R.sup.2]=.47, [R.sup.2] change=.06, P [is less than or equal to] .05.

Logistic regression using PPT scores to predict level of care of subjects showed that this variable alone correctly predicted 72 subjects' (90%) living situations. Three subjects were predicted to be living in a dependent care setting, although they actually lived in an independent setting. Five subjects were identified as independent, although they lived in a dependent care setting.

Discussion

The extremity musculoskeletal measures examined in this study account for a substantial amount of the variation in functional ability of older adults. This finding is consistent with prior research that has shown that musculoskeletal impairments are related to function[12,14,15] and disability.[3,4,8-11] In our study, age, lower-extremity ROM, and lower-extremity muscle force were predictors of functional ability, as measured by the PPT, for a general sample of older adults with various levels of functioning and independence.

We used cross-sectional data, so it is impossible to determine whether musculoskeletal impairment causes lower performance on functional tasks. An alternative explanation for these findings is that muscle force and ROM decline due to disuse when individuals experience functional decline.

When analyzed according to subjects' level of care, the extremity impairments predicting function differed for independent living elderly subjects as compared with those living in dependent care environments. Upper-extremity ROM was an important predictor of function for the independent group, but it did not contribute to function for the dependent living group. Upper-extremity muscle force and lower-extremity ROM were important in determining function for dependent subjects, but not for independent subjects.

In addition to the different impairments identified for the 2 subgroups, the amount of variability in function scores accounted for by musculoskeletal factors differed for dependent living subjects as compared with independent living subjects. For the dependent group, age and musculoskeletal factors accounted for 68% of the variability in function scores, whereas these variables explained only 47% of the variability for the independent sample. The subjects in the independent living settings had low levels of musculoskeletal impairments when compared with those in the assisted living settings. It is likely that performance on the PPT failed to challenge many of the independent living subjects to their musculoskeletal limit. Eleven of the 42 subjects in the independent living group scored 100% on the PPT, raising the concern of a ceiling effect for this group. For this study, we used the 7-item version of the PPT, excluding items related to stair climbing. Inclusion of the stair climbing items may have been more challenging for the independent subjects, reducing the possibility of the ceiling effect. However, concern for the safety of the most frail subjects and a desire to use the same instrument for the entire study sample prevented inclusion of these items.

A modified version of the PPT has been developed and is proposed as being more appropriate for relatively high-functioning older adults.[28] This modified PPT includes higher-order tasks, such as repeated rising from a chair and a tandem standing task, as substitutes for 2 fine motor tasks in the original version. Use of the modified PPT, or another instrument that measures higher-order tasks, could reduce the ceiling effect for the relatively high-functioning older adults, but it may be less appropriate to use in frail elderly adults.

An initial concern with the use of the PPT as the measure of outcome was that relatively few studies have examined the validity of data obtained with the instrument. Reuben and colleagues[20,22] established concurrent validity for the PPT in comparing PPT scores with standardized measures of basic ADL and IADL tasks and elderly subjects' self-reports of functional abilities. Other researchers[24,28] have used the instrument to categorize elderly individuals as "fit" or having "mild" or "moderate" frailty or as being unable to function in the community without assistance. Our study provides support of the predictive validity of data obtained with the PPT, as these scores correctly classified the level of care received by 90% of the subjects. Assuming that all subjects participating in the study were correctly placed in dependent versus independent living situations, our results suggest that the PPT could be a useful tool for deciding whether an individual can live independently or requires a higher level of care.

A limitation of the study was the focus on only extremity measures of musculoskeletal function. To minimize the burden of testing on subjects, we did not include measures of all extremity joints in terms of muscle force and ROM, but rather selected those used in prior related research.[25] Therefore, not all measures of interest were included.

We chose to focus our study on the effects of extremity impairments on function in older adults and, therefore, did not consider other impairments that logically may affect physical function in this population. For example, impairments in spinal mobility or force may be important factors limiting function in older adults. Schenkman and colleagues[29] found that spinal mobility was associated with performance of functional tasks such as reaching, turning 360 degrees, and moving from a supine position to a sitting position. It is also likely that nonmusculoskeletal factors such as reaction time, cardiopulmonary condition, and cognitive status are predictive of functional performance in older adults.

Although we examined the relationships between musculoskeletal impairments and a global measure of function, other researchers[13-15,30] have examined the effects of specific musculoskeletal force or ROM measures on discrete measures of function (eg, ability to rise from a chair, speed of rising from a chair) and gait variables. These prior studies provide important information about how impaired muscle force, for example, relates to specific functional limitations.

The information from our study can be used to design exercise programs for older adults with the goal of restoring or maintaining function. From our results, it would appear that exercises focusing on lower-extremity ROM and upper-extremity muscle force could improve function in older adults in dependent living situations, whereas older adults living independently would receive the most gains in function from participating in exercise interventions targeting upper-extremity ROM. These outcomes, however, can only be determined from a prospective study. Although general interventions have been shown to be effective in restoring function to older adults,[31] further research is needed to determine whether interventions such as improving muscle force or ROM in the extremities directly affects function.

Summary and Conclusions

The data from this study support the use of the PPT as a predictor of disability in older adults, as PPT scores alone accurately predicted the living situation of 90% of the subjects in the study. Extremity muscle force and ROM are predictors of function in older adults, explaining a substantial amount of the variance in function scores. Overall, lower-extremity muscle force and ROM make the primary contributions to function. However, different musculoskeletal factors relate to function for individuals in dependent living settings as compared with individuals in independent living settings. Upper extremity muscle force and lower-extremity ROM are important predictors of function for elder people living in assistive living or skilled nursing facilities, whereas upper-extremity muscle force is the most critical predictor for elderly people living independently in the community. Future research should focus on the effects of specific exercise interventions on function in older adults.

(*) JTech Medical Industries, 357 West 910 South, Heber City, UT 84032.

([dagger]) SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611.

References

[1] Gill TM, Williams CS, Tinetti ME. Assessing risk for the onset of functional dependence among older adults: the role of physical performance. J Am Geriatr Soc. 1995;43:603-609.

[2] Leon J, Lair T. Functional Status of the Noninstitutionalized Elderly: Estimates of ADL and IADL Difficulties--National Medical Expenditure Survey Research Findings 4. Rockville, Md: US Dept of Health and Human Services, Public Health Service; 1990. Agency for Health Care Policy and Research Publication (PHS) 90-3462.

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

[4] Jette AM, Assmann SF, Rooks D, et al. Interrelationships among disablement concepts. J Gerontol A Biol Sci Med Sci. 1998;53:M395-M404.

[5] Fried LP, Guralnik JM. Disability in older adults: evidence regarding significance, etiology, and risk. J Am Geriatr Soc. 1997;45:92-100.

[6] Pope AM, Tarlov AR, eds. Disability in America: Toward a National Agenda for Prevention. Washington, DC: National Academy Press; 1991.

[7] Jette AM, Branch LG. Impairment and disability in the aged. J Chronic Dis. 1985;38:59-65.

[8] Jette AM, Branch LG, Berlin J. Musculoskeletal impairments and physical disablement among the aged. J Gerontol. 1990;45:M203-M208.

[9] Guralnik JM, Ferrucci L, Simonsick EM, et al. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332:556-561.

[10] Hughes SL, Dunlop D, Edelman P, et al. Impact of joint impairment on longitudinal disability in elderly persons. J Gerontol. 1994;49:S291-S300.

[11] Lawrence RH, Jette AM. Disentangling the disablement process. J Gerontol B Psychol Sci Soc Sci. 1996;51:S173-S182.

[12] Bergstrom G, Aniansson A, Bjelle A, et al. Functional consequences of joint impairment at age 79. Scand J Rehabil Med. 1985;17:183-190.

[13] Woolley SM, Topp RV, Khuder SA, et al. Function: which factors predict ability in OA patients? Biomechanics in Geriatrics. December 1998:6-12, 17.

[14] Hughes MA, Myers BS, Schenkman ML. The role of strength in rising from a chair in the functionally impaired elderly. J Biomech. 1996;29:1509-1513.

[15] Bassey EJ, Bendall MJ, Pearson M. Muscle strength in triceps surae and objectively measured customary walking activity in men and women over 65 years of age. Clin Sci (Colch). 1988;74:85-89.

[16] 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 measures. J Am Geriatr Soc. 1992;40:457-462.

[17] Fried LP, Storer DJ, King DE, Lodder F. Diagnosis of illness presentations in the elderly. J Am Geriatr Soc. 1991;39:117-123.

[18] Rubenstein LZ, Schairer C, Wieland GD, Kane R. Systematic biases in functional status assessment of elderly adults: effects of different data sources. J Gerontol. 1984;39:686-691.

[19] Karagiozis H, Gray S, Sacco J, et al. The Direct Assessment of Functional Abilities (DAFA): a comparison to an indirect measure of instrumental activities of daily living. Gerontologist. 1998;38:113-121.

[20] 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.

[21] Glass TA. Conjugating the "tenses" of function: discordance among hypothetical, experimental, and enacted function in older adults. Gerontologist. 1998;38:101-112.

[22] 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.

[23] 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.

[24] Crews L, Brown M, Norton BJ, et al. Effectiveness of the Physical Performance Test for Detecting and Monitoring Changes in Functional Status in Elderly Patients Receiving Home Physical Therapy Services [thesis]. St Louis, Mo: Washington University; 1997.

[25] Gerety MB, Mulrow CD, Tuley MR, et al. Development and validation of a physical performance instrument for the functionally impaired elderly: the Physical Disability Index (PDI). J Gerontol. 1993;48:M33-M38.

[26] Andrews AW, Thomas MW, Bohannon RW. Normative values for isometric muscle force measurements obtained with hand-held dynamometers. Phys Ther. 1996;76:248-259.

[27] Joint Motion: Method of Measuring and Recording. Chicago, Ill: American Academy of Orthopaedic Surgeons; 1965.

[28] Brown M. The physical performance test for the assessment of frailty. GeriNotes. 1998;5(4):7-11.

[29] Schenkman M, Shipp KM, Chandler J, et al. Relationships between mobility of axial structures and physical performance. Phys Ther. 1996;76:276-285.

[30] Brown M, Sinacore DR, Host HH. The relationship of strength to function in the older adult. J Gerontol A Biol Sci Med Sci. 1995;50:55-59.

[31] Landefeld CS, Palmer RM, Kresevic DM, et al. A randomized trial of care in a hospital medical unit especially designed to improve the functional outcomes of acutely ill older patients. N Engl J Med. 1995;332:1338-1344.

KL Beissner, PT, PhD, is Associate Professor and Chair, Department of Physical Therapy, Ithaca College, 335 Smiddy Hall, Ithaca, NY 14850 (USA) ([e-mail]). Address all correspondence to Dr Beissner.

JE Collins, PT, MPA, is a doctoral candidate in the Margaret Warner Graduate School of Education and Human Development, University of Rochester, Rochester, NY. She was Assistant Professor, Department of Physical Therapy, Ithaca College, at the time this study was conducted.

H Holmes, MA, is Project Manager of the Pathways to Life Quality Study, Gerontology Institute, Ithaca College.

All authors provided writing and consultation (including review of manuscript before submission). Dr Beissner and Ms Collins provided data collection, and Dr Beissner and Ms Holmes provided data analysis. Dr Beissner provided concept/research design, project management, and fund procurement. Ms Collins provided subjects and institutional liaisons. The authors acknowledge the students who assisted in the study, especially Sara Alima, Kristen Kara, and Marie Ticen.

A portion of this research was presented at the Combined Sections Meeting of the American Physical Therapy Association, February 11-15, 1998, Boston, Mass, and at the New York Physical Therapy Association Annual Meeting, October 8-10, 1999, Rye Brook, NY.

This research was supported by the New York Physical Therapy Association Research Designated Fund and by the Ithaca College Gerontology Institute.

This study was approved by the All-College Review Board for Human Subjects Research at Ithaca College and by the Human Subjects Review Board at St John's Home, Rochester, NY.

This article was submitted December 22, 1998, and was accepted February 16, 2000.