Measures of Physical Ability Are Unrelated to Objectively Measured Physical Activity Behavior in Older Adults Residing in Continuing Care Retirement Communities

Article Outline

• Abstract
• Methods
  • Statistical Analysis
• Results
• Discussion
  • Study Limitations
• Conclusions
• References
• Copyright

Abstract 

Zalewski KR, Smith JC, Malzahn J, VanHart M, O'Connell D. Measures of physical ability are unrelated to objectively measured physical activity behavior in older adults residing in continuing care retirement communities.

Objective

To explore the relationship between measures of physical performance, physical activity, and self-reported physical activity.

Design

Cross-sectional analysis.

Setting

Continuing care retirement communities (CCRCs) in the greater Milwaukee area.

Participants

Older adults from independent or assisted living apartments (N=59).

Interventions

Not applicable.

Main Outcome Measures

Physical performance was measured with comfortable and fast gait speeds and the six-minute walk test. Physical activity was measured by an ankle-mounted accelerometer to observe daily steps; stepping rate was averaged over a 7-day wear time. Self-reported physical activity was measured by the Physical Activity Scale for the Elderly (PASE).

Results

Participants reported walking more steps per day than older adults who are not living in CCRCs. There was no relationship between physical abilities and total steps walked (r=.087–.213, P>.05). No relationship was observed between measures of physical performance or total steps and PASE scores (r=−.034–.177, P = not significant). The relative contributions of physical activity categories to total PASE score were different than published reports for older adults not living in CCRCs.

Conclusions

Common measures of physical performance often used by clinicians in making decisions on rehabilitation outcomes do not appear to be related to the actual functioning of older adults residing in senior communities. The nature of the environment customized to the needs of the older adult may facilitate increased physical activity participation independent of physical abilities.

Key Words: Aged, Rehabilitation, Residence characteristics

List of Abbreviations: CCRCs, continuing care retirement communities, MMSE, Mini-Mental State Examination, PASE, Physical Activity Scale for the Elderly, 6MWT, six-minute walk test, SAM, StepWatch Activity Monitor

CLINICIANS USE MEASURES of physical performance collected in the clinic environment to inform predictive judgments of physical performance in the natural environment, to make decisions on discharge criteria, and to assess the continued need for therapy intervention. Comfortable gait speed, fast gait speed, and the 6MWT are common outcome measures tracked in rehabilitation environments, and normative values for age and sex have been published.¹ These tests have been found to be sensitive to change in healthy older adults² and in adults with specific medical diagnoses.³, ⁴ Although gait speeds have been shown to be predictive of discharge placement⁴ and functional ambulation for people with stroke,⁵ the broader relationship of walking ability measured in the clinic to ability witnessed in the community is not known.

The few studies that do explore the relationship between measures of physical performance in clinic and field environments report a lack of reliable association between the two. Moseley and colleagues⁶ reported poor agreement between clinic measures of gait speed and endurance and covert field measures, including walking in parking lots and shopping centers, in persons with traumatic brain injury (intraclass correlation coefficient_1,1, range .06–.29 for comfortable gait speed, fast gait speed, and 6MWT). Field studies exploring physical activity in older adults that use activity monitors generally find that older adults walk fewer steps per day and engage in more sedentary behavior (ie, fewer bouts of activity) than younger people.⁷, ⁸, ⁹

To determine whether activity monitor data are consistent with subjective reports of activity behavior of older adults, Washburn and Flicker¹⁰ examined healthy older adults who wore an activity monitor for 3 days and then completed the PASE. The PASE is a survey designed to describe and categorize various daily activity behaviors (eg, light housekeeping vs yard work), which are weighted to reflect energy expenditure. The findings suggested statistically significant but only moderate relationships (r=.49–.64 depending on age) between measured steps per day and total PASE scores. The PASE was found to be moderately correlated with physical abilities as measured by the 6MWT (r=.35) in a group of older adults with knee pain due to osteoarthritis.¹¹

There is little information on the physical performance and daily activity of adults living in CCRCs. These CCRCs are designed to allow older adults to age in a community of other older adults. Typically, these environments differ from other retirement communities because they offer people a range of living support, from independent apartments to assisted living and skilled nursing care. The level of care chosen by a resident is often specified in the initial contract and may be renegotiated as needs change. A CCRC has the advantage of allowing people to age in a community even if the level of care needs to be changed. As these CCRCs develop, they may become powerful mediators in moderating the health behaviors of their residents.

The purposes of this study are 2-fold. First, we aim to describe the physical activity characteristics of older adults residing in CCRCs. Second, we explore the relationships between 3 measures of physical performance, a field measure of physical activity, and a subjective report of physical activity in older adults residing in CCRCs.

Back to Article Outline

Methods 

We report data collected as part of a larger study¹² exploring physical activity, physical abilities, and psychosocial factors in older adults. Approval for the study was obtained from the Institutional Review Board of the University of Wisconsin, Milwaukee. The study was conducted in a set of 5 CCRCs, providing services to adults age 55 and older. The CCRCs included in the request to participate in the study included 10 facilities known to the principal investigator. Five of the facilities contacted agreed to participate in the research. These facilities represent 3 private nonprofit religion-affiliated institutions and 2 private for-profit facilities. All of the facilities were within a 60-mile radius of the University of Wisconsin, Milwaukee. This study included only people who resided in independent or assisted living apartments in the participating CCRCs. Eligible participants were required to self-report walking as their primary means of mobility inside the facility but were allowed to use an assistive device. Participants were excluded in this analysis if their MMSE¹³ score was below 21 of 30. Participants were recruited through informational sessions (total attendance, 81 residents), fliers distributed to the residents' mailboxes (approximate distribution, 400 fliers), and through word of mouth. The 5 facilities participating in the study represent approximately 400 apartments (some apartments had more than 1 resident). Sixty participants were recruited from these 5 facilities.

Participants completed 1 session of data collection. After providing informed consent, participants completed a battery of tests. The tests were administered in the following order: MMSE,¹³ PASE,¹⁴ comfortable gait speed, fast gait speed, and the 6MWT. Gait speeds and the 6MWT were completed following the protocol as outlined in Steffen et al.¹ Participants completed 2 trials of walking over a 10-m distance; the middle 8 m were collected and analyzed for gait speeds. The 6MWT was completed on an inside path within the resident facility. Although paths were constrained by the environment, they all had a minimal distance of at least 40m and were oval or round. Participants were encouraged at minute 3 and minute 5 with reports of remaining time and “good work” feedback. Heart rate and blood pressure were recorded before and after the 6MWT; a rating of perceived exertion was collected only after the 6MWT with the 6 to 20 Borg rating of perceived exertion scale.¹⁵ Immediately after the walk, participants were asked to rate their fatigue on a 0 to 10 scale, with 0 described as “no fatigue at all—like sitting in a chair” and 10 described as “the most fatigue you have ever experienced.”

At the end of the session, each participant was fitted with a SAM,^a which is attached to the ankle with a small Velcro strap. The SAM is a dual-axis accelerometer designed to count steps in people with atypical gait characteristics, including those with abnormal or slow gaits, and can be specifically configured to take into account individual gait characteristics. For the current study, all participants used the same settings: the quick stepping option was set to no, walking speed and leg motion were set to normal, and range of speeds was set to uses a moderate range of speeds. Height settings were customized to each participant. Participants wore the SAM for 7 consecutive days. They were instructed to apply the SAM upon waking and to only remove the SAM if it became uncomfortable or interfered with sleep. Because the SAM is waterproof, participants were encouraged to wear the SAM during water-based physical activity (eg, water aerobics). Data collected from the 7-day period included total steps, highest mean step rate for 60 continuous minutes, and highest mean step rate for 20 continuous minutes. The latter 2 estimates represent the best continuous minutes of activity during the respective time window.

The PASE questionnaire comprises descriptions of activities typically engaged in by people who reside in their own homes. Participants reported their current level of participation in sitting, walking, light recreation, moderate recreation, strenuous recreation, and strength training activities (never, seldom [1–2d/wk], sometimes [3–4d/wk], or often [5–7d/wk]), and how many hours per day on average were spent engaged in these activities. Participants also reported household and work/volunteer-related activity (yes or no). The total PASE score was computed by multiplying the amount of time spent in each activity by item weights.¹⁴, ¹⁶

Statistical Analysis 

Means and standard deviations were calculated for all variables. Relationships between comfortable and fast gait speeds, 6MWT, total PASE score, and total steps were explored by the Pearson product-moment correlation. Individual-level differences in comfortable and fast gait speeds, and stepping rates over short (20min) and longer (60min) performances were evaluated by paired t tests.

Back to Article Outline

Results 

Among those recruited, 59 participated (mean age, 83.8y, 45 women). Four of the participants were recruited from assisted living apartments; the remainder of the participants resided in independent apartments. Additional demographic data are provided in table 1.

Table 1. Demographic Data for Study Participants

Variables	Frequency (%)
Resident facility (n=59)
1	40
2	50
3	7
4	0⁎
5	2
Marital status (n=56)
Married	28
Widowed	33
Divorced	11
Single	8
Highest achieved education (n=59)
Elementary	0
High school	52
College	32
Graduate	15
Number of physician visits annually (n=59)
0	2
1	20
2	33
3 or more	44
Number of prescription medications taken daily (n=59)
0	17
1–2	27
3–4	30
5 or more	24
Self-reported comorbidities (% reporting “yes”) (n=59)
Cardiovascular	71
Psychiatric	8
Medical/metabolic	70
Orthopedic	64

NOTE. Percentages have been rounded to the nearest whole number.

Although no participants were ineligible because of low MMSE scores, not all participants completed the entire test battery. (The entire test battery required 2h to administer.) If participants were unable to contribute that amount of time, they were allowed to withdraw from the study; these participants allowed their collected data to be used for analysis. One participant withdrew from the study early in the data collection process, stating that the data collection was “too cumbersome,” and refused to allow any data to be used for analysis. Results of tests of physical performance and PASE scores are presented in table 2. Physical abilities of study participants were below published age-related normative values.¹ Participants in this study had higher reported average total steps per day than reported in other literature.⁷ In an attempt to evaluate variability in daily walking behavior within the day, mean stepping rate for the best continuous 60 minutes and best continuous 20 minutes of activity during the 7-day testing window are reported, as are comfortable and fast gait speeds. Although there was a difference between comfortable and fast gait speeds (t₅₇=−9.126, P<.05), there was no difference in step rate measured during a longer (60min) compared with a shorter (20min) walking duration (t₅₂=.229, P=.82).

Table 2. Performance on Tests of Physical Ability and Daily Ambulatory Activity

Abbreviations: CV, coefficient of variation; NA, not applicable.

There was no relationship between tests of physical ability, physical activity, and total PASE scores (table 3).

Table 3. Relationships Between Physical Abilities, Physical Activity, and Energy Expenditure

Variables	6MWT	CGS	FGS	TSPD	PASE
TSPD					.073(N=47)
FGS				.087(N=51)	.177(N=48)
CGS			.812⁎(N=57)	.123(N=51)	.173(N=48)
6MWT		.649⁎(N=53)	.639⁎(N=53)	.213(N=47)	−.034(N=44)

NOTE. Pearson product-moment correlation (N=47–56).

Abbreviations: CGS, comfortable gait speed (m/s); FGS, fast gait speed (m/s); TSPD, total steps per day (no. of steps).

Total PASE score distribution is presented in table 4. For this sample, 64.5% of the total PASE score was due to exercise activity, 5.1% was from work or volunteer activities, and 30.4% was due to daily household related activity. These relative contributions are considerably different than those reported in the literature.¹⁴, ¹⁶

Table 4. Contribution to Total PASE Score by PASE Component

Back to Article Outline

Discussion 

This study did not find a relationship between measures of physical performance, physical activity, and PASE scores. One explanation for this finding may be that older adults residing in CCRCs do not engage in behaviors that contribute to the total PASE score in the same proportion as do other independently living adults. Specifically, light housework, heavy housework, outdoor gardening, home repair, and lawn and yard care, which contributed 60% of the total PASE score in a study of independent older adults living in their homes,¹⁴ contributed only 30.4% of the total PASE score of older adults in CCRCs. Typically, much of the more strenuous home care is provided by the facility and thus would not be reported as physical activity behavior by the PASE. This explanation is supported with an analysis of the relationship between the best continuous minutes of activity and PASE subscale scores. The best continuous 60 minutes of physical activity was correlated with the subcomponent of walking (Pearson r=.314, P<.05), and moderate activity (Pearson r=.281, P<.05); however, the best 20 minutes of physical activity did not correlate significantly with any of the PASE subscale measures.

Measures of daily physical activity suggest that people living in CCRCs may actually be more active than independently living seniors not residing in senior-designed environments. Considering that the participants in this study appeared more physically limited on the basis of measures of physical performance compared with independently living older adults, the high daily step count in the current sample suggests that the residential environment may be encouraging behaviors that minimize the impact of physical limitations on daily physical activity. Noting that measures of physical performance were not related to daily steps further supports the suggestion that aspects of the residential environment may minimize physical limitations. However, this finding must be interpreted with caution. The SAM may overcount actual activity by approximately 3% of total steps,¹⁷ leading to an overestimation of actual activity. Although the older adults in this study appeared very similar to other published reports that used the SAM as a measurement tool,⁷ there is the possibility that these step counts are artificially high.

Another explanation may be that participants in this sample have overreported moderate and strenuous activity in the PASE. It is possible that the participants in this study found the rating of activity intensity to be ambiguous, perhaps reflecting a difficulty separating the construct of the “difficulty” of a behavior from its “aerobic intensity.” Although we did not directly test this hypothesis in our study, it is suggested by comparing the coefficient of variation of the 6MWT walk distance (28%) against the coefficient of variation of the 6MWT perceived exertion (82%). The perception of effort was highly variable in this sample, despite a fairly consistent walking distance. Individual differences in cardiovascular fitness may also explain the high variability in perceived exertion. Although fitness was not directly assessed in this study, the coefficient of variation for heart rate after the 6MWT was also low (29%). Finally, it is possible that the PASE is not valid in this population.

The lack of relationship between physical performance and steps walked per day is more difficult to explain. It may be that people engage in many short bouts of walking that add up to high daily activity scores, and are less constrained by walking speed or endurance in CCRC environments. Most facilities are designed to facilitate this type of walking behavior, encouraging short walks to dinner, to get mail, and to most central congregation areas. This possibility is further supported by the lack of difference between stepping rates for the best continuous 60 minutes and best continuous 20 minutes of walking. Participants in this sample were consistent in their daily activity behaviors.

Study Limitations 

Results of this study should be interpreted cautiously. It is possible that only the healthiest and most active people from the CCRCs volunteered for the study, introducing a source of potential bias. Additionally, most of the participants were from 2 of the 5 CCRCs. These 2 facilities represent large facilities in the greater Milwaukee area. It is possible that each facility has a set of cohort characteristics that may not be reflective of residents of CCRCs in general.

Because therapists often have used measures of physical performance, particularly comfortable gait speeds, to make clinical decisions regarding patient abilities outside the rehabilitation environment, the lack of relationship between gait speeds and 6MWT and measures of physical activity is important. This study suggests that therapists should be cautious in predicting participation in routine daily physical activity that is based on common clinical measures, and that they should consider the residential living environment when interpreting test scores. The SAM provides an easy method for objective and accurate measurement of daily physical activity in a home environment and should be considered as a complementary tool to assess outcomes.

Back to Article Outline

Conclusions 

The 6MWT, and comfortable and fast gait speeds as measures of physical performance are not good predictors of objectively measured daily physical activity behavior among residents of continuing care rehabilitation communities.

Supplier

Back to Article Outline

References 

1. Steffen T, Hacker T, Mollinger L. Age- and gender-related performance in community-dwelling elderly people: Six-Minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Phys Ther. 2002;82:128–137
   • View In Article
   • MEDLINE
2. Perera S, Mody S, Woodman R, Studenski S. Meaningful change and responsiveness in common physical performance measures in older adults. J Am Geriatr Soc. 2006;54:743–749
   • View In Article
   • MEDLINE
   • CrossRef
3. English C, Hillier S, Stiller K, Warden-Flood A. The sensitivity of three commonly used outcome measures to detect change amongst patients receiving inpatient rehabilitation following stroke. Clin Rehabil. 2006;20:52–55
   • View In Article
   • MEDLINE
   • CrossRef
4. Salbach N, Mayo N, Higgins J, Ahmed S, Finch L, Richards C. Responsiveness and predictability of gait speed and other disability. Arch Phys Med Rehabil. 2001;82:1204–1212
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (162 KB)
   • CrossRef
5. Kollen B, Kwakkel G, Lindeman E. Time dependency of walking classification in stroke. Phys Ther. 2006;86:618–625
   • View In Article
   • MEDLINE
6. Moseley A, Lanzarone S, Bosman J, et al. Ecological validity of walking speed assessment after traumatic brain injury. J Head Trauma Rehabil. 2004;19:341–348
   • View In Article
   • MEDLINE
   • CrossRef
7. Cavanaugh J, Coleman K, Faines J, Laing L, Morey M. Using step activity monitoring to characterize ambulatory activity in community dwelling older adults. J Am Geriatr Soc. 2007;55:120–124
   • View In Article
   • MEDLINE
   • CrossRef
8. Davis M, Fox K. Physical activity patterns assessed by accelerometry in older people. Eur J Appl Physiol. 2007;100:581–589
   • View In Article
   • MEDLINE
   • CrossRef
9. Matthews C, Chen K, Freedson P, et al. Amount of time spent in sedentary behaviors in the United States. Am J Epidemiol. 2008;167:875–881
   • View In Article
   • CrossRef
10. Washburn R, Flicker J. Physical Activity Scale for the Elderly (PASE): the relationship with activity measured by a portable accelerometer. J Sports Med Phys Fitness. 1999;39:336–340
    • View In Article
    • MEDLINE
11. Martin K, Rejeski J, Miller M, James K, Ettinger W, Messier S. Validation of the PASE in older adults with knee pain and physical disability. Med Sci Sports Exerc. 1999;31:627–633
    • View In Article
    • MEDLINE
    • CrossRef
12. Smith J, Zalewski K, Motl R, O'Connell D, Malzahn J. Self-efficacy and physical activity behavior among elders in an assisted living environment. Med Sci Sports Exerc. 2008;40(5 Suppl 1):S468
    • View In Article
    • CrossRef
13. Hartford Institute for Geriatric Nursing. “Mini-Mental State.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–198
    • View In Article
    • Full-Text PDF (635 KB)
    • CrossRef
14. Washburn R, Smith K, Jette A, Janney C. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol. 1993;46:153–162
    • View In Article
    • MEDLINE
    • CrossRef
15. Borg G. Borg's perceived exertion pain scales. Champaign (IL): Human Kinetics; 1998;
    • View In Article
16. Schuit A, Schouten EG, Westerterp K, Saris W. Validity of the Physical Activity Scale for the Elderly (PASE): according to energy expenditure assessed by the doubly labeled water method. J Clin Epidemiol. 1997;50:541–546
    • View In Article
    • Abstract
    • Full-Text PDF (637 KB)
    • CrossRef
17. Storti K, Pettee K, Brach J, Talkowski J, Richardson C, Kriska A. Gait speed and step-count monitor accuracy in community-dwelling older adults. Med Sci Sports Exerc. 2008;40:59–64
    • View In Article
    • CrossRef

• a StepWatch Activity Monitor (SAM); OrthoCare Innovations, 700 12th St SW, Ste 700, Washington, DC 20005.