Decline in Physical Performance Among Women With a Recent Transient Ischemic Attack or Ischemic Stroke
Opportunities for Functional Preservation A Report of The Women’s Estrogen Stroke Trial
Background and Purpose— Physical performance for walking, reaching, turning, and other common tasks is a major determinant of functional independence after stroke. Current strategies to preserve physical performance focus on prevention of recurrent stroke. Loss of physical performance, however, may occur in the absence of recurrence. To examine this possibility, we measured change in physical performance, independent of subsequent stroke, among women with a recent ischemic stroke or transient ischemic attack (TIA).
Methods— Among 664 postmenopausal women who participated in a clinical trial of estrogen therapy after stroke or TIA, we administered the Physical Performance Test (PPT) at baseline (mean 58 days from the cerebrovascular event) and annually. Women who died or had a stroke during follow-up were censored. Decline or improvement in physical performance was defined as a change in the PPT score from baseline of at least 3 points. Sustained decline or improvement was defined as 2 consecutive years during which the score had declined or improved, respectively, relative to the baseline score.
Results— With each year of follow-up, a smaller proportion of the cohort demonstrated improvement (16% in year 1, 6% in year 5) and a larger proportion demonstrated decline (15% in year 1, 35% in year 5). In an analysis restricted to 259 women with 3 years of follow-up, 46 (18%) experienced a nonsustained decline in physical performance, and 39 (15%) experienced a sustained decline.
Conclusions— Decline in physical performance is common after an ischemic stroke or TIA even in the absence of a recurrent neurological event. Our findings suggest that specific interventions to maintain and improve physical performance may be important for reducing long-term disability.
Among older persons, physical performance for walking, turning, reaching, hand actions, and other common tasks is a major determinant of functional independence (ie, independence for basic and instrumental activities of daily living such as dressing, eating, toileting, shopping, transportation, and housekeeping).1–3 In recognition of this relationship, stroke physicians commonly recommend physical therapy to maximize physical performance in the period immediately after a stroke. To maintain physical performance for the long-term, stroke physicians recommend therapies and behavioral changes to prevent stroke recurrence. Although prevention strategies are justified by the high rate and neurological consequence of stroke recurrence, they do not account for the possibility that long-term decline in physical performance may occur independent of recurrence.
The potential importance of mechanisms unrelated to recurrence is suggested by research among community-living older persons showing that ≈10% lose functional independence annually in one or more basic activities of daily living.1 Approximately half of all patients who lose this independence do so in the absence of a major illness requiring hospitalization.4 In response to this observation, investigators have recently demonstrated that simple interventions to maintain physical performance can prevent loss of independence among community-living older persons.5 These findings may have particular applicability to stroke patients who, because of baseline neurological impairment, may be at high risk for declining physical performance that can precipitate functional dependence. To our knowledge, however, there have been no investigations to document risk for declining performance after recovery from an acute stroke.
The purpose of our research, therefore, was to determine the incidence and risk factors for decline in physical performance, independent of subsequent stroke, among women with a recent transient ischemic attack (TIA) or ischemic stroke. Our hypothesis was that decline would be common and that important risk factors would include poor baseline functional and cognitive status.
Subjects and Methods
A complete description of the design and methods for the Women’s Estrogen for Stroke Trial (WEST) has been published.6 The WEST was a double-blind, randomized, placebo-controlled trial to examine the effectiveness of estradiol for prevention of recurrent stroke and death among postmenopausal women with a recent TIA or ischemic stroke. Potential participants were identified from admissions to 21 hospitals in Connecticut and Southern Massachusetts. Postmenopausal women older than age 44 years were eligible for enrollment if they had a nondisabling ischemic stroke or TIA (index event) within the previous 90 days. To be considered nondisabled, participants had to score <5 on the NASCET severity scale7 and be functionally independent (ie, able to eat, dress, and walk a few steps without assistance).
Women were not eligible if they had a contraindication to estrogen therapy, a severe comorbid condition that limited life-expectancy to <5 years, a severe neurological or psychiatric disease that would complicate endpoint evaluation, or if they were using estrogen at the time of the index event.
Among 664 randomized participants in the WEST, we excluded from the analysis 31 who did not have a baseline assessment of physical performance and 76 who did not have a physical performance score at 1 year because of death (n=28) or interim stroke (n=48). The final cohort comprised 557 WEST participants. The WEST was approved by the institutional review boards of all participating hospitals.
Before randomization, we interviewed and examined all participants in their home. The mean time from the qualifying cerebrovascular event to interview was 57.9 days (range, 6 to 134). Cognitive function was assessed with the Mini Mental State Examination (MMSE).8,9 The Centers for Epidemiological Research Depression scale was used to assess depressive symptoms.10 Neurological impairment was assessed using the National Institutes of Health Stroke Scale,11 which is a measure of residual deficits for 11 categories of neurological signs. Self-reported functional status in 10 activities of daily living was recorded using the Barthel Index.12
The Physical Performance Test (PPT) measures performance on 7 tasks, including writing a sentence, simulated eating, lifting a book and putting it on a shelf, putting on and removing a jacket, bending over and picking up a penny, circling, and walking for 25 feet.13 For each task, participants receive a score based on time to completion. Scoring for simulated eating, for example, is as follows: ≤10 seconds=4 points, 10.5 to 15 seconds=3 points, 15.5 to 20 seconds=2 points, >20 seconds=1 point, and not able to complete=0 points. The maximum possible score for the seven tasks is 28 points. Scores were considered missing if participants were unable to complete a task because of nonphysical impediments (eg, refusal, no room to perform test, poor eyesight, illiteracy).
Follow-up Procedures and Outcome Assessment
Research nurses visited participants in their homes annually to re-administer the PPT. If a participant experienced a stroke during follow-up, they were withdrawn from the trial.
For the current analysis, the primary outcome was a clinically meaningful decline in physical performance, which was defined as a decrease from baseline of 3 points in the PPT score. This criterion was derived from an analysis of the total PPT score at baseline for the entire cohort (mean, 17.6; standard deviation, 5.8). Three points was selected because it represents 50% of the standard deviation of the baseline score, which is a generally accepted indicator of moderate effect size.14 This 3-point criterion for clinically significant change in PPT score is slightly more conservative than a 2-point change, which has been advocated by other investigators.15 To more fully describe the experience of the cohort, we also examined improvement in physical performance, defined as a 3-point increase in PPT score. Sustained decline or improvement in physical performance, as opposed to a change lasting only 1 year, was defined as 2 consecutive years during which the PPT score was decreased or increased at least 3 points relative to the baseline score.
To describe change in physical performance among WEST participants, we used 2 strategies. First, we described year-to-year change in PPT score for mutually exclusive subgroups defined by total duration of follow-up in the trial. That is, year-to-year change in score (ie, increase, decrease, or no change relative to baseline) was described separately for participants with 1, 2, 3, 4, and 5 years of follow-up. Second, we categorized WEST participants with at least 3 years of follow-up according to 5 trajectories of physical performance (ie, sustained improvement, nonsustained improvement, no change, nonsustained decline, sustained decline).
To identify risk factors for sustained decline in physical performance, time to sustained decline (defined as days from baseline PPT to first annual test that met the criterion for decline and which was followed by a subsequent, consecutive measurement that met the criterion for decline from baseline) was analyzed among participants with known PPT trajectories over the first 3 years using proportional hazards modeling. The comparison group for this analysis comprised participants with known PPT trajectories over the first 3 years of trial participation who had no decline from baseline in any year. A forward selection algorithm, with probability value criteria for entry of 0.10, was used to identify participant characteristics that were predictive of sustained decline. All baseline features with sufficient prevalence in the cohort in the 3-year cohort (n >10) were considered for selection.16 Continuous measures (ie, education, Centers for Epidemiological Research Depression scale, MMSE, National Institutes of Health Stroke Severity score, Barthel score) were evaluated for inclusion simultaneously in their original and dichotomized forms. For dichotomous scores, we selected conventional values demarcating depression on Centers for Epidemiological Research Depression scale (≥16),17 cognitive impairment on MMSE (<24),9 any disability on Barthel (<100), and any neurological impairment on the National Institutes of Health Stroke Severity score (>1). Because we were interested in identifying features predictive of sustained decline after controlling for initial level of physical performance, we included baseline PPT score in each model. We also estimated the final multivariable model among participants who scored in the top tertile of PPT at baseline to further assess the effect of the selected features in this high-functioning subgroup.
Trajectories of Physical Performance
Baseline features for the 557 participants are displayed in Table 1. The mean duration of follow-up was 2.7 years, with a range of <1 year to 5 years. Year-to-year change in physical performance, accounting for variable durations of follow-up, are displayed in Table 2. As shown in the top row, with each year of follow-up a smaller proportion of the cohort demonstrated improvement (16% in year 1, 6% in year 5) and a larger proportion demonstrated decline (15% in year 1, 35% in year 5). Not shown in this Table is an analysis among all 557 participants showing that 192 (34%) of women experienced a decline in physical performance, relative to baseline, at some point during follow-up. Among 259 participants with complete PPT information over the first 3 years of follow-up, 25 participants (10%) experienced a sustained improvement, 99 participants (39%) had no change in PPT score over time (ie, in any year, the PPT score did not deviate from baseline by >±3 points), and 39 (15%) experienced a sustained decline. Ninety-six participants (37%) experienced a nonsustained improvement or decline.
Risk Factors for Sustained Decline in Physical Performance
Features associated with increased risk for sustained decline in physical performance included older age, lower cognitive function (ie, lower MMSE score), history of previous stroke, stroke at entry, and lower Barthel score (Table 3). As an alternative strategy to control for the effect of baseline physical performance score, we restricted the analysis to the 82 participants in the top tertile of baseline PPT score. The relative risk estimates for sustained decline for the selected features were similar in direction and magnitude to the larger sample.
To understand which components of physical performance contributed to the overall decline in PPT score, we compared change in component scores from baseline to year 3 among participants who experienced sustained decline to change among participants in the comparison group who experienced no decline (Table 4). Among the 7 performance tasks, the greatest decline in score among sustained decliners (mean decline >1.0 points) was observed for bending over to pick up an object from the floor, circling, and dressing.
The main finding from this research is that decline in physical performance, independent of recurrent stroke, is common among postmenopausal women in the years immediately after a TIA or ischemic stroke. In our cohort of 557 women, 192 (34%) of women experienced a decline in physical performance at some point during 1 to 5 years of follow-up. In each successive year after the TIA or stroke, furthermore, a larger proportion of participants performed at a level substantially below their baseline. Among women with at least 3 years of follow-up, 39 (15%) experienced a sustained decline. Based on limited data from other studies, the rate of decline we observed among women with a recent stroke is probably larger than rates observed among men and women assembled in community surveys.18,19 Differences in definitions of decline, however, limit direct comparison between our work and these studies.
Risk factors for sustained decline in our cohort included older age, lower cognitive function, history of previous stroke, stroke at entry (compared with TIA), and poor baseline functional status (according to the Barthel score). These findings are consistent with research on community-living older persons.1,2,20 In that geriatric research, however, the outcome has usually been loss of functional independence (eg, disability in one or more activity of daily living), not decline in physical performance, as in our research. Despite these differences in functional outcomes, our results and those of previous population-based studies of older persons demonstrate the predictive importance of baseline functional and cognitive status.
A second important finding in our research is that physical performance is a dynamic characteristic of persons with symptomatic cerebrovascular disease. In women with known trajectories over 3 years, 96 (37%) experienced temporary (nonsustained) improvements or declines in physical performance. Similar findings for activities of daily living have been reported for community-living elderly persons.21 The finding of such fluctuation demonstrates that improvement is possible and decline is not always permanent.
Because physical performance is a powerful determinant of functional independence, a major implication of our research is that prevention of decline in physical performance may be an important, distinct objective for care after stroke. Aspirin, carotid endarterectomy, anticoagulation, and hypertension control are currently the main therapeutic strategies after stroke, and all are designed to prevent recurrence. None of these interventions, however, will preserve functional independence in the majority of patients who do not experience a recurrent stroke. Among all patients with stroke, therefore, interventions to preserve physical performance may emerge as critical adjunctive therapies for achieving the ultimate goal of preserving functional independence. Preliminary research among “at-risk” community-living older persons, furthermore, suggests that effective interventions targeting impairments in physical abilities can be conducted using a simple home-based protocol.5 In essence, our findings indicate the potential effectiveness of modern geriatric therapeutics for patients with cerebrovascular disease. With 15% of cerebrovascular disease patients demonstrating a sustained decline in physical performance within 3 years of their event, the absolute number of persons who might benefit from this approach is large.
The PPT used for this research was first described in 1990 as an objective measure of physical performance for older persons.13 Although it has excellent reliability and validity,13,15,22 investigators have since developed timed tests that may be more sensitive to change and more closely associated with risk for functional dependence. These include standing balance, chair stand, and integrative tests such as the “get up and go test.”1,2,23,24 The timed rapid gait and circling tasks, which are included in the PPT, are still widely used but the full PPT itself is not. These newer tests emphasize balance and lower extremity and large muscle group performance, which, in our own data, were more sensitive to change over time than measures of upper extremity performance. It is possible that these newer tests may more accurately identify decline in physical performance among patients who have recovered from cerebrovascular disease.
In summary, among patients with nondisabling ischemic cerebrovascular events, sustained decline and fluctuation in physical performance is common. If confirmed in other stroke populations, and with other measures of physical performance, these findings could warrant testing of strategies to prevent, detect, and remediate impairments in physical performance. Maintaining physical performance may represent a novel approach to preserving functional independence in this vulnerable population.
Funding provided by The National Institute of Neurological Disorders and Stroke, National Institutes of Health (grant number: 1-RO1-N531251); John R. Marler, MD, Medical Officer. Additional support and study drug was provided by Mead Johnson Laboratories. T.M.G. is the recipient of a Midcareer Investigator Award in Patient-Oriented Research (K24AG021507) from the National Institute on Aging.
- Received November 10, 2004.
- Accepted November 22, 2004.
Gill TM, Williams CS, Tinetti ME. The combined effects of baseline vulnerability and acute hospital events on the development of functional dependence among community-living older persons. J Gerentol. 1999; 54A: M377–M383.
Anonymous. Special report from the National Institute of Neurological Disorders and Stroke. Classification of cerebrovascular diseases III. Stroke. 1990; 21: 637–676.
Radloff LS. The CES-D Scale. A self-report depression scale for research in the general population. Appl Psychol Meas. 1977; 1: 385–401.
Brott T, Adams HP, Olinger CP, et al. Measurement of acute cerebral infarction: a clinical examination scale. Stroke. 1989; 20: 864–870.
Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Maryland State Med J. 1965: 61–65.
Reuben DB, Siu AL. An objective measure of physical function of elderly outpatients. The Physical Performance Test. JAGS. 1990; 38: 1105–1112.
King MB, Judge JO, Whipple R, Wolfson L. Reliability and responsiveness of two physical performance measures examined in the context of a functional training intervention. Physical Therapy. 2000; 80: 8–16.
Peduzzi PN, Concato J, Holford TR, Feinstein AR. The importance of events per variable in multivariable analysis. II. Accuracy and precision of regression estimates J Clin Epidemiol. 1995; 48: 1503–1510.
Bjornsbo KS, Ferry M, De Groot CPGM, Schlienger JL. Changes in physical performance in elderly europeans. Seneca 1993–1999. J Nutrition Health Aging. 2002; 6: 9–14.
Gill TM, Williams CS, Richardson ED, Tinetti ME. Impairments in physical performance and cognitive status as predisposing factors for functional dependence among nondisabled older persons. J Gerentol. 1996; 51A: M283–M288.
Gill TM, Kurland B. The burden and patterns of disability in activities of daily living among community-living older persons. J Gerentol. 2003; 58A: 70–75.
Reuben DB, Siu AL, Kimpau S. The predictive validity of self-report and performance-based measures of function and health. J Gerentol. 1992; 47: M106–M110.