| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Reports |
AK Welmer, PT, MSc, is a doctoral student in the Division of Physiotherapy, Department of Neurobiology Care Sciences and Society, Karolinska Institutet, SE-141 83 Huddinge, Sweden, and the Vårdal Institute, Lund, Sweden
M von Arbin, MD, PhD, is Senior Consultant and Head, Stroke Unit, Division of Medicine, Department of Clinical Sciences, Karolinska Institutet Danderyd Hospital, Stockholm, Sweden
V Murray, MD, PhD, is Clinical Scientist and Consultant, Division of Medicine, Department of Clinical Sciences, Karolinska Institutet Danderyd Hospital
L Widén Holmqvist, PT, PhD, is Associate Professor, Division of Physiotherapy, Department of Neurobiology Care Sciences and Society, Karolinska Institutet
DK Sommerfeld, PT, PhD, is Senior Physiotherapist, Department of Geriatric Medicine, Danderyd Hospital, Danderyd, Sweden, and University Lecturer, Division of Physiotherapy, Department of Neurobiology Care Sciences and Society, Karolinska Institutet
Address all correspondence to Ms Welmer at: anna-karin.welmer{at}ki.se
Submitted November 17, 2006;
Accepted July 26, 2007
 | Abstract |
|---|
Subjects and Methods: In 115 subjects who were 
65 years of age, the following parameters were assessed 5 days after stroke: somatosensory (touch and proprioceptive), perceptual, and cognitive functions; depressive symptoms; mobility; and self-care.
Results: Multivariate analyses showed that normal proprioceptive function was significantly associated with better mobility. Normal perceptual and touch functions were significantly associated with better self-care. Subjects with normal proprioceptive function were 8.6 times as likely to have normal perceptual function as subjects with proprioceptive impairment.
Discussion and Conclusion: Somatosensory and perceptual functions were significantly associated with subjects' activity levels. Normal proprioceptive function also might indicate normal perceptual function.
| Introduction |
|---|
 TopAbstract Introduction MethodResultsDiscussionConclusionReferences |
|---|
It has been shown that normal somatosensory (touch and proprioceptive) function is associated with high activity levels 10 days after acute stroke.6 However, impaired somatosensory function is an uncertain indicator of negative outcomes.6
Studies investigating the association between impaired perceptual functions (here defined as the mental functions of recognizing and interpreting sensory stimuli) or impaired cognitive functions and activity limitations after stroke have shown conflicting results.7,8 The relationship between perceptual or cognitive impairments and activity limitations during the first week after stroke has not been well described so far. Depression after acute stroke has been reported to be associated with activity limitations for a multitude of theoretical reasons, many of which have not yet been clearly defined.9,10
Perceptual function may be difficult to assess at an early stage after stroke because it demands a patient's attention. Therefore, perceptual impairments may be overlooked at this point after stroke. Considering the potential importance of perceptual function, a test indicating whether perceptual function is impaired or not is needed.
The present study focused on older people with stroke (65 years of age), who represent a rapidly expanding part of the population. The rationale for focusing on older people is that functioning among them is not as well described as it is among younger people. Furthermore, impairments and activity limitations may differ between younger and older people.11 However, it has been suggested that rehabilitation is effective for very old people as well.12
One aim of the present study was to describe somatosensory (touch and proprioceptive), perceptual, and cognitive functions as well as depressive symptoms and their association with mobility and self-care 5 days after acute stroke in older people (65 years of age). Another aim was to investigate whether the assessment of somatosensory functions also will provide information on perceptual functions in that population.
| Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
A total of 115 subjects (52 men and 63 women), with a mean age of 81 years (range=65–96), were included according to the inclusion and exclusion criteria. Twenty-three subjects had had a previous stroke. Fifty-six subjects had right-side movement-related symptoms, 54 had left-side symptoms, 1 had bilateral symptoms, and 4 had no such symptoms but did have activity limitations.
The subjects were examined for functioning as described below at 5±1 (
±SD) days after stroke. When possible, the subjects were examined while resting. The examinations were conducted by physical therapists and occupational therapists who were cotrained for the purpose.
Assessments of Somatosensory Functions
Touch function was determined by testing the ability to perceive pinprick (metal pin) and light touch (cotton, wool) on the upper arm, forearm, dorsal surface of the hand, thigh, calf, and dorsal surface of the foot. If a subject was unable to perceive pinprick or light touch in one location or more on the affected side, then the test result was defined as impaired touch function. These tests are frequently used and are considered to show satisfactory reliability, with a rough grading into normal or impaired touch function.13
The proprioceptive function of the affected upper limb was tested with the Thumb Localizing Test.14 The upper limb of the affected side was positioned passively, and the subject was asked to pinch the thumb of that limb with the opposite thumb and index finger 4 times. Proprioceptive function was considered to be normal if the subject was able to locate the thumb on the affected side in 3 of the 4 tests with his or her eyes closed. The test is considered to be valid14 but has not been tested for reliability.
Assessments of Perceptual Functions
Perceptual space function was assessed with the Line Cancellation Task15 and the Letter Cancellation Task.16 The subjects were allowed unlimited time to complete the tests. The score was the number of targets canceled. In the Line Cancellation Task, 4 lines in the center of a sheet of paper were used for demonstration, leaving 36 lines to cancel. The subjects were asked to cancel the remaining lines, which were ruled in a standard fashion on the sheet of paper. The suggested normal value is 32 to 36 canceled lines.17 In the Letter Cancellation Task, the subjects were asked to cancel 29 of 30 target "A"s among 120 distracting letters displayed at random on a sheet of paper. One "A" in the center of the sheet of paper was used for demonstration. No normal value has been established. In the present study, we used the same percentage definition of normal as in the Line Cancellation Task, that is, 26 to 29 canceled targets. These cancellation tests are widely used and are considered to be valid tests of perceptual space function. At least 2 different cancellation tests are recommended to identify all subjects with impaired attention.18
Perceptual functions were further evaluated with respect to color sense and visuospatial construction by use of the Kohs Block Design Test.19 There are 16 cubes with red, white, blue, yellow, red-white, and blue-yellow sides (one of each). The colors on the 2-colored sides are divided diagonally. Four cubes are used in tests 1 to 4, 9 cubes are used in tests 5 and 6, and 16 cubes are used in test 7. One point is given for each construction. The test leader demonstrated the task by building 2 constructions from given designs presented on pictures. The subjects were allowed unlimited time to complete the test. This block test is considered to be reliable20 and valid21 in evaluating everyday spatial function.
Assessment of Cognitive Functions
Cognitive functions were screened with the Mini-Mental State Examination (MMSE).22 A normal range of values has been established at 24 to 30 points.23 The MMSE is considered to be reliable24 and to show acceptable validity in detecting cognitive dysfunction at an early stage after stroke in older people.25 Subjects with interrupting aphasia or dysarthria were not assessed with the MMSE in the present study.
Assessment of Depressive Symptoms
Selected depressive symptoms were assessed with the following items of the Montgomery Åsberg Depression Rating Scale (MADRS)26: item 2, reported sadness; item 4, reduced sleep; item 5, reduced appetite; and item 7, lassitude. The MADRS is considered to be reliable and valid for individual items.26
Assessment of Activity Levels
Mobility and self-care were assessed with the Rivermead Mobility Index (RMI)27 and the Barthel Index (BI).28 Subjects with less than 4 points on the RMI and subjects with less than 35 points on the BI after acute stroke are considered to be severely disabled.4 The RMI and the BI are considered to be reliable and valid.27–29
Basic Demographic Data and Lateralization of the Brain Lesion
The age, sex, occurrence of previous stroke, and brain lesion side of the subjects were recorded. The clinical scales, the ranges of the scores, the cutoff scores for disability, and the normal scores are shown in Table 1.
|
Logistic regression analyses were conducted to examine possible associations between somatosensory (touch and proprioceptive), perceptual, and cognitive functions and depressive symptoms as independent variables and mobility and self-care as dependent variables, with age, sex, previous stroke, and brain lesion side as covariates. Univariate models were used to assess the unadjusted odds ratios (ORs) and 95% confidence intervals (CIs). Multivariate models (forward stepwise) were used to assess the adjusted ORs and 95% CIs. An OR above 1 indicates an increased chance of achieving a better activity score, and a value below 1 indicates a decreased chance. The adjusted OR is an estimate of the relative chance of achieving a better activity score when all other independent variables are adjusted for. Forward stepwise selection combines the backward elimination and forward selection methods to add variables to the model or to remove variables from the model as they meet or as they fail to meet specified significance levels, respectively. All variables with a significance level of P<.1 were included in the multivariate models. The P value for entering the forward stepwise multivariate logistic regression analyses was set at .1. The sensitivity of the models was derived from the overall correct predictions at a predicted probability level of .5. Using the Mann-Whitney U test, we compared the mobility and self-care scores for subjects who were able to complete the body function tests with those of subjects who were not able to do so. The Bonferroni correction was used to correct for multiple comparisons.
In order to establish whether the assessment of somatosensory functions also will provide information on perceptual functions, likelihood ratios (LRs) for positive values were calculated. The LR for a positive result is the ratio of the chance of normal perceptual function if a subject has normal somatosensory function to the chance of normal perceptual function if the subject does not have normal somatosensory function. The significance level was set at P<.05. Data were analyzed with STATISTICA 7.0 for Windows.*
| Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
|
|
Multivariate logistic regression analyses showed that normal proprioception, as determined with the Thumb Localizing Test, was significantly associated with better mobility according to the RMI (OR=3.4, 95% CI=1.1–10.6). Normal perceptual function, as determined with the Kohs Block Design Test, tended to be significantly associated with better mobility according to the RMI (P=.08) (OR=3.3, 95% CI=0.9–12.7). The sensitivity of the model, that is, subjects with a true low mobility level, was 39%. The multivariate analyses showed that normal perceptual function, as determined with the Kohs Block Design Test (OR=7.3, 95% CI=1.8–29.0), and normal touch function in the lower extremities (OR=3.9, 95% CI=1.0–15.3) were independently associated with a better self-care score according to the BI. Not having had an earlier stroke tended to be significantly associated with a better self-care score according to the BI (P=.06) (OR=4.5, 95% CI=0.9–22.3). The sensitivity of the model was 58%. No interaction effects were found.
Table 4 shows the LRs for positive results of the somatosensory tests in determining the status of the perceptual tests. The highest LR (8.6) was seen for the Thumb Localizing Test in indicating the outcome of the Kohs Block Design Test.
|
| Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
Normal somatosensory function was significantly associated with better mobility and self-care. The multivariate analyses showed that the chance for people with normal proprioceptive function, relative to people with impaired proprioceptive function, to achieve better mobility was 3.4. The chance for people with normal touch function in the lower extremities to achieve better self-care was 3.9. Somatosensory function, especially proprioceptive feedback, has been shown to play an important role in the coordination of movements.31 Our results are in accordance with those of Desrosiers et al,32 who suggested that lower-extremity impairments are associated with activity limitations more than upper-extremity impairments are, perhaps because it is easier to compensate for impairments in the upper extremities than for those in the lower extremities.
We found no relationship between cognitive functions, as assessed with the MMSE, and activity levels. Our result agrees with the findings of Hajek et al,8 who reported no association between those variables in the first week after stroke. Indeed, it has been suggested that some basic activities, such as walking and stair climbing, are unaffected by cognitive dysfunction, in contrast to more complex activities.33 Another interpretation of our findings is that the assessments of mobility and self-care may have been constructed in a way that underestimates the importance of cognitive functions required for activity tasks. However, because some subjects could not be assessed with the MMSE, mainly because of an inability to follow instructions, it is possible that these subjects had poorer cognitive functions than the subjects who could be assessed. The subjects who could not be assessed showed poorer activity scores than those who could be assessed. Therefore, an ultimate conclusion that no association exists between cognitive functions and activity levels in older people with stroke cannot be drawn. Further studies are needed to explore whether or to what extent cognitive functions affect mobility and self-care in people with stroke or whether these activities can be improved separately and independently of cognitive functions. Ruchinskas et al33 found that cognitive functions, including perceptual function, were associated with self-care but not with mobility in older people with mixed diagnoses in a rehabilitation unit. However, those findings are not fully comparable to ours because we separated perceptual and cognitive functions.
The univariate analyses in the present study showed that perceptual function was significantly associated with self-care and mobility (Tab. 3). Furthermore, the multivariate analyses showed that the chance for people with normal perceptual function to achieve better self-care was 7.3 and that the chance for such people to achieve better mobility was 3.3, although the latter value was not quite statistically significant (P=.08). The difficulties for people with perceptual impairments in performing mobility and self-care activities may result from a failure to take into account the aspects of the situation necessary for the performance of complex sequences of actions, such as eating and dressing. Our results confirm those of Paolucci et al,7 who reported that perceptual function was more strongly associated with self-care than with mobility in people assessed 5 weeks after stroke. Considering our results and those of Paolucci et al, it is reasonable to assume that mobility is a basic behavior to a greater degree than self-care is.
The present study showed that activity levels were more strongly associated with visuospatial construction, as assessed with the Kohs Block Design Test, than with perceptual space function, as assessed with the cancellation tasks, after acute stroke. The reason for these findings may be that visuospatial construction requires not only perceptual space function but also flexibility in the use of spatial reference systems.34
The results of the present study showed that people with normal proprioceptive function, as assessed with the Thumb Localizing Test, were 8.6 times as likely to have normal perceptual function, as assessed with the Kohs Block Design Test, as people with impaired proprioceptive function. The Thumb Localizing Test was chosen because it is easy for both the examiner and the subject to carry out. It is closely linked, although not equal, to tests of joint position and movement.14 The test examines the position of a passively fixed thumb in relation to the body axis, as measured with motor tasks performed by the contralateral reaching limb.14
Of the selected items from the MADRS, item 2 (reported sadness) was chosen because it represents a core depressive symptom, equivalent to the first of the 2 cardinal symptoms in the diagnostic criteria for depression listed in the Diagnostic and Statistical Manual of Mental Disorders: DSM-IV.35 The remaining items were chosen in order to depict the autonomic situation and because they had the advantage of being reportable by subjects or observable by relatives or staff. It has been suggested that the severity of depressive symptoms is related to activity limitations.9 We found that, of the selected items, only the true depressive symptom, that is, reported sadness, was significantly associated with dependence with regard to self-care in the univariate analyses. A lower score, that is, less pronounced sadness, on MADRS item 2 increased the chance of achieving a better self-care score by 1.4. This finding confirms that of Ramasubbu et al,10 who reported that global self-rated depression was related to dependence with regard to self-care.
Although the association between sadness and self-care in the present study was not found to be significant in the multivariate analyses, it nevertheless seems important to assess sadness because it indicates that depression should be assessed and that the diagnosis should be verified or ruled out according to DSM-IV diagnostic criteria. If a person is found to be depressed, treatment might positively influence that person's activity levels. Some items of the MADRS were not considered to have a marked impact on activity levels and therefore were not included in the present study, and this decision might have affected the results. However, the inclusion of the most prominent depressive symptom, sadness, lends credibility to the importance of the findings for depression in the univariate analyses.
The subjects who could not be examined with any of the tests for somatosensory, perceptual, and cognitive functions or for depressive symptoms formed a homogeneous group, in which the inability to be assessed acted as a common indicator of low mobility and self-care scores. The fact that these subjects were not included in the analyses skews the results toward the subjects who were less impaired and limits the generalization of our results. However, more than half of the patients had already been discharged from the hospital at 5 days after stroke.
The limitations of our study are that the cross-sectional data did not allow us to examine the cause-effect relationship between the body functions and activities assessed. In addition, the 95% CIs were generally wide, presenting an element of uncertainty in the estimates. Further studies with larger samples or systematic reviews with meta-analyses are needed to establish the exact influence of somatosensory and mental functions on activity levels after acute stroke.
The cutoff values chosen for the different scales may have affected our results. This factor is always a potential cause of varying study results and constitutes a cause of difficulties in comparisons of studies. There is a lack of normative values and consensus in the literature for cutoff values for the Letter Cancellation Task and the Kohs Block Design Test. Therefore, the cutoff for the Letter Cancellation Task was estimated on the basis of that of the Line Cancellation Task. For the Kohs Block Design Test, the cutoff of 0/>0 seemed most plausible in order to minimize the skewed distribution toward the subjects who were less impaired because so many subjects could not be assessed with the test. Furthermore, a dichotomous outcome was chosen for the somatosensory tests because it is considered to be more reliable than a graded scale.13 The cutoff values used for the RMI and BI were chosen because an RMI score of less than 4 and a BI score of less than 35 shortly after acute stroke have been found to have a negative influence on the recovery of activity levels, discharge destination, and length of hospital stay.4,5
| Conclusion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
| Footnotes |
|---|
This study was approved by the Ethics Committee of Karolinska University Hospital.
This study was supported by grants from the Swedish Stroke Association, the Solstickan Foundation, the Vårdal Institute, and Karolinska Institutet, Stockholm, Sweden (to Ms Welmer); from the Stockholm County Council (to Dr Sommerfeld); from the Marianne and Marcus Wallenberg Foundation (to Dr Murray); and from AFA Insurances (to Dr Murray).
* StatSoft Scandinavia AB, Sportfältsvägen 3, SE-754 19 Uppsala, Sweden. 
| References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
