The importance of physical activity to health is well established. After stroke, regular physical activity is critical for regulating blood glucose and promoting decreases in body weight, blood pressure, level of total blood cholesterol, serum triglycerides, and low-density lipoprotein cholesterol [
Walking performance has been found to be significantly associated with free-living physical activity in community-dwelling stroke survivors [
However, less attention has been paid to the influence of impairments, that is, a loss or abnormality of body structure and function [
Personal factors (such as age, BMI, psychological characteristics, and social support) may influence physical activity after stroke. Although physical activity has not been investigated directly, the presence of a spouse at home and good social support have been found to predict the ability to carry out activities of daily living in the long term [
Given that some impairments and personal factors are amenable to modification in people after stroke, an understanding of which of these are associated with free-living physical activity may assist in planning appropriately targeted interventions. Therefore, the aim of this study was to determine which impairments and/or personal factors are most associated with free-living physical activity in community-dwelling people after stroke.
A cross-sectional observational study was carried out with community-dwelling people after stroke. Ambulatory people with chronic stroke were recruited from the local community within a major city. Personal factors and impairments were collected on one day, and free-living physical activity was collected over two days in the community. Each participant was randomly allocated a day of the week and wore the activity monitor on this day across two consecutive weeks. The days for measurement of free-living physical activity were counterbalanced across the week so that there was the same amount of data collected for each day of the week. Data were collected from 30 min after getting out of bed (i.e., after dressing) until 30 min prior to going to bed (i.e., before undressing). Participants were instructed to carry out their routine activities. All measurements for each participant were completed within a 2-week period.
People with stroke were included if they were within 1 to 5 years of their first stroke, over 50 years old, and able to walk 10 m independently without an aid. They were excluded if they could not speak English or if they were unable to follow instructions. Ethical approval was obtained from the Human Research Ethics Committee at the local institution. Informed consent was obtained from all participants before data collection commenced.
Age, gender, weight, height, side of hemiplegia, time since stroke, and presence of spouse were collected. Weight and height were used to calculate BMI in kg/m2.
Eleven impairments were measured and were divided into two categories: sensory-motor (including muscle weakness, contracture, spasticity, loss of coordination, proprioception, and balance) and non-sensory-motor (including cognitive, language, perceptual, mood abnormalities, and loss of confidence). Measures were chosen on the basis that they were easy and quick to perform in the clinic (e.g., they did not require extensive equipment), that they measured the impairment directly (i.e., they were not a subsection of a larger scale), and where possible, that they were valid and reliable for use in neurological conditions or with elderly patients. One person took all measures in order to eliminate inter rater variability and where relevant; the affected leg was measured.
Strength of the knee extensors was measured using hand-held dynamometry [
Contracture of the plantarflexors was measured using the method of Moseley and Adams [
Spasticity of the plantarflexors was measured using the Tardieu scale [
Coordination of the lower limb was measured using the Lower Extremity Motor Coordination Test (LEMOCOT) [
Proprioception of the knee joint was measured using a matching task [
Balance was measured using a modified version of the Single Leg Stance Test [
In terms of non-sensory-motor impairments, cognition was measured using the Mini-Mental State Examination (MMSE) [
Language was measured using the Frenchay Aphasia Screening Test (FAST) [
Perception was measured as neglect using the Line Bisection Test adapted from Olk and Harvey [
Mood was measured using the 6-item self-report Short Depression-Happiness Scale (SDHS) [
Confidence was measured using the 10-item self-report General Perceived Self-efficacy Scale [
Free-living physical activity was collected using an activity monitor-Intelligent Device for Energy Expenditure and Activity (IDEEA). This device is light (58 g), and the recorder is clipped to the belt or waist of the pants. It monitors body motion through five sensors attached to the front of the chest, to the front of both thighs, and underneath both feet using medical tape. Postures (lying, reclining, sitting, standing, and leaning), transitions (lie to sit, sit to lie, recline to sit, sit to recline, recline to stand, stand to recline, sit to stand, and stand to sit), and gait (walking, running, up and down stairs, and jumping on both legs) are measured. An investigator visited participants’ homes and calibrated the device. The recording of physical activity was then begun, with the investigator returning to turn the device off and check the data at the end of the day. Free-living physical activity was reported as duration (time on feet and time not on feet) and frequency of activity (activity counts) carried out per person per day [
The IDEEA has been found to be >98% accurate for duration, frequency, type, and intensity of a variety of physical activities in normal adults [
We collected data on 42 participants so that if up to 8 variables were entered into the regression analysis, there would be at least 5 cases per independent variable [
Shapiro-Wilk normality test was used to determine if the free-living physical activity data was normally distributed. It showed that the variable “activity counts” was positively skewed. When a log transformation was performed which normalized the activity counts data, there was no difference in the regression results. That is, the deviation from normal was not so large as to affect the outcome of the analysis. Therefore, the original data was used in order to facilitate interpretation.
Univariate analysis was undertaken using Pearson’s correlation coefficient to examine the association between characteristics and free-living physical activity. Characteristics with correlations of
Forty-two stroke survivors aged 70 years (SD 10) with a BMI on the upper limit of normal participated in this study (Table
Characteristics of participants.
Characteristic | Participants ( |
---|---|
Personal factors | |
Age (yr), mean (SD) | 70 (10) |
Gender, | 29 (69) |
Weight (kg), mean (SD) | 73 (12) |
Height (m), mean (SD) | 1.7 (0.1) |
Side of hemiplegia, | 23 (55) |
Time since stroke (yr), mean (SD) | 2.8 (1.4) |
BMI (kg/m2), mean (SD) | 26.4 (4.3) |
Living with spouse, | 37 (88) |
Impairments | |
Sensory-motor impairments, mean (SD, range) | |
Strength ( | 116 (52, 53–303) |
Contracture (°) | 4.8 (4.8, 0–19) |
Spasticity (0 to 4) | 1.2 (1.0, 0–3) |
Coordination (taps/s) | 0.6 (0.4, 0–1.6) |
Proprioception (°) | 4.2 (2.6, 0–12.6) |
Balance (s) | 5 (7, 0–30) |
Non-sensory-motor impairments, mean (SD, range) | |
Mood (0 to 18) | 11 (4, 3–18) |
Confidence (10 to 40) | 29 (5, 17–39) |
Perception (mm) | 2.2 (1.6, 0–5) |
Cognition (0 to 30) | 25 (3, 17–30) |
Language (0 to 20) | 17 (4, 2–20) |
Free-living physical activity over waking day, mean (SD, range) | |
Time on feet (min) | 230 (115, 29–506) |
Activity counts (#) | 5656 (4091, 543–18804) |
The mood score of 18 participants (43%) was below 10 which fulfills the criteria for depression. Most of the participants were confident and most also bisected the horizontal line within 5 mm of the middle, suggesting good perception. On average, the cognitive score was within normal limits; however, the scores of 11 (27%) participants were <24 suggesting the presence of cognitive impairment in some participants. Moreover, the Frenchay Aphasia Screening showed that 8 (19%) participants had language impairment.
The mean duration of free-living activity monitored was 10.8 hr/day (SD 1.3). On average, participants spent 230 min (SD 115) on their feet which was 35% of the monitored time. On average, they registered 5656 activity counts (SD 4091).
Univariate analysis showed that balance and mood were significantly correlated with time on feet (
Univariate analysis of the correlation between characteristics and free-living physical activity using Pearson’s correlation coefficient
Characteristic | Free-living physical activity | |
Time on feet | Activity counts | |
Personal factors | ||
Age | −0.18 (0.26) | −0.26 (0.09) |
Gender | 0.10 (0.52) | 0.17 (0.28) |
Side of hemiplegia | 0.12 (0.45) | 0.00 (0.99) |
Time since stroke | 0.00 (0.99) | −0.03 (0.85) |
BMI | −0.29 (0.06) | −0.12 (0.45) |
Living with spouse | 0.00 (0.98) | 0.12 (0.44) |
Impairments | ||
Sensory-motor impairments | ||
Strength | 0.18 (0.25) | 0.03 (0.85) |
Contracture | −0.28 (0.07) | −0.27 (0.09) |
Spasticity | −0.15 (0.33) | −0.21 (0.18) |
Dexterity | 0.10 (0.52) | 0.15 (0.33) |
Proprioception | 0.07 (0.68) | −0.02 (0.89) |
Balance | 0.42 (<0.01) | 0.54 (<0.001) |
Non-sensory-motor impairments | ||
Mood | 0.43 (<0.01) | 0.52 (<0.001) |
Confidence | −0.03 (0.84) | 0.09 (0.57) |
Perception | −0.13 (0.42) | −0.26 (0.10) |
Cognition | 0.18 (0.25) | 0.03 (0.85) |
Language | 0.11 (0.50) | −0.04 (0.80) |
When the characteristics that were correlated with time on feet (
Similarly, when the characteristics that were correlated with activity counts (
The aim of this study was to determine which characteristics were most associated with free-living physical activity in community-dwelling people after stroke who could walk independently. Balance and mood were associated with free-living physical activity in community-dwelling stroke survivors, regardless of whether physical activity was measured as time being active or frequency of activity.
There was only one sensory-motor impairment—balance—that was associated with free-living physical activity such that the poorer the balance, the lower the amount of physical activity. Even though our sample could all walk 10 m unaided, standing on one leg was less than one quarter of normal performance in our participants [
There was also only one non-sensory-motor impairment—mood—that was associated with free-living physical activity such that the greater the depression, the lower the amount of physical activity. 42% of our population met the criteria of depression, which is similar to findings from two community-based studies of poststroke depression [
The finding that balance and mood are associated with physical activity does not mean that depression and poor balance cause low levels of physical activity—the relationship is more likely to be cyclical. For example, if stroke survivors have poor balance, their physical activity is likely to be curtailed, which in turn may lead to a worsening in balance from lack of practice, thereby setting up a vicious cycle of deterioration in balance and physical activity. Similarly, depression can also curtail physical activity, which in turn may lead to further depression [
In our study, spouse support was not correlated with free-living physical activity, which differs from Jorgensen and colleagues’ [
There are several implications from the findings of this study for clinicians involved in rehabilitation after stroke. Our findings suggest that intervention aimed at improving balance and enhancing mood may be useful in promoting long-term physical activity. Although the evidence for the benefit of exercise in managing depression is not clear-cut in the nonstroke population [
The strengths of this study were that an activity monitor was used to measure free-living physical activity by which the limitations of self-report methods such as recall bias was avoided [
We have found that balance and mood were associated with free-living physical activity in community-dwelling people after stroke who can walk independently. These findings provide guidance to professionals working in rehabilitation in appropriately targeting intervention. Further research is required to examine prospectively whether early intervention to improve mood and balance will increase free-living physical activity of this population.
The authors certify that no party having a direct interest in the results of the research supporting this paper has or will confer a benefit on them or on any organization with which we are associated.
The authors would like to thank the University of Dammam in Saudi Arabia for funding Matar Alzahrani’s doctoral studies and Gemma Lloyd for her help in recruiting participants.