Paretic upper-limb strength best explains arm activity in people with stroke.

Stroke is one of the leading causes of disability in the older population and can significantly affect aspects of a person's physical, emotional, and social life. As stroke mortality rates decline, (1) individuals are more likely to have residual impairments that could affect daily living. More than 80% of individuals with stroke experience hemiparesis, (2,3) and of those people who initially have upper-extremity paresis, it is estimated that 70% have residual impairment. (4,5) The upper limb makes a significant contribution to most activities of dally living (ADL), and impairments can compromise participation in many of these essential and meaningful tasks. The return of upper-limb function has been identified as an important rehabilitation goal. (6) Consequently, knowledge of upper-limb impairment and its relationships to activity (eg, performance of daily tasks) and participation is necessary in order for clinicians to plan effective and efficient rehabilitation.

Upper-limb impairments following stroke can include weakness, pain, sensory loss, impaired dexterity, and incoordination. Recently, Desrosiers and colleagues (7) demonstrated a significant relationship between upper-limb motor impairment using the Fugl-Meyer Motor Impairment Scale and upper-limb function measured with the Functional Independence Measure. Lai and colleagues (8) used the National Institutes of Health Stroke Scale to demonstrate the predictive nature of upper-limb motor impairment on Barthel Index (BI) scores, and Nakayama and colleagues (9) used the Scandinavian Stroke Scale to determine the effect of upper-limb impairment on upper-limb function (using the BI). However, these studies used global measures of upper-limb impairment (ie, combined impairments into one score) that do not assess the individual contribution of specific impairment variables such as strength (force-generating capacity), tone (resistance to passive movement), and sensation in determining upper-limb function. In addition, they used ADL measures that do not focus exclusively on upper-limb performance. For example, the BI is heavily weighted on general mobility functions, and a high score (indicating independence) can be achieved without adequate recovery of the paretic upper limb). (10) Quantifying the contribution of specific impairments to upper-limb function could assist clinicians in treatment planning during rehabilitation.

Relatively few studies have explored the relationship between specific upper-limb impairments, such as altered tone (11) and muscle weakness, (11-13) and upper-limb function during ADL. In previous research, weakness was correlated with poor performance in hand-to-mouth action (r=.83) (11) and ADL performance as measured with the Upper Extremity Performance Test for the Elderly (r = .63-.88), (12,13) whereas tone (measured with the Modified Ashworth Scale) did not correlate with functional movement. (11) Only one study (14) assessed the contribution of several impairment predictors (motor recovery, tone, and sensation) to establish the most important factor of arm function. Although Fugl-Meyer Motor Impairment Scale scores were found to be the most significant predictor of upper-limb function using the BI, the measures used were not specific enough to target the individual impairments associated with poor upper-limb function.

The International Classification of Functioning, Disability and Health (15) (ICF) has 3 domains: bodily function and structures, activity, and participation. Difficulties occurring in the domain of bodily function and structures are called "impairments." We based our impairment variables on specific impairments that are commonly assessed and treated in clinical practice, and quantifying their contribution to upper-limb performance could assist clinicians in treatment planning during rehabilitation. We chose to assess upper-limb strength, tone, sensation, and pain. The choice of these 4 impairment variables was based on their evaluative prevalence in studies of upper-limb function after stroke. (7,11,12,14)

The ICF defines activity as the execution of a task or action by an individual, (15) and activity limitations are difficulties that an individual may have in performing activities. We selected the subset of ADL from the wider domain of activity. Activities of daily living are considered essential to independent living and includes activities such as dressing, eating, and carrying. The Chedoke Arm and Hand Activity Inventory (CAHAI) (16) and the Motor Activity Log (MAL) (17) are measures of upper-limb performance in ADL and were chosen to reflect this important aspect of activity. The CAHAI is a measure of upper-limb capacity where 13 ADL items are scored based on performance of the paretic upper limb. The MAL is a self-report measure that evaluates real-world use of the paretic upper limb in ADL outside of an experimental setting (ie, in the home and community).

Participation is a relatively new focus embraced by the ICF model, (15) although it was acknowledged under the term "handicap" in the previous International Classification of Impairments, Disabilities, and Handicaps (ICIDH) model, (18) Society-perceived participation defines involvement in life roles that are typical for someone of similar age, sex, or background. (19) Although such measures are useful for between-group comparisons, person-perceived participation (an individual's perception of his or her involvement in life situations) may be more clinically relevant in meeting an individual's needs and for attaining information that may lead to the development of relevant interventions. (19) The Reintegration to Normal Living (RNL) Index (20) was developed for the purpose of evaluating how people with chronic conditions (eg, stroke) regard their involvement in self-care, recreational, and social activities. We chose this measure because of its ability to reflect the concept of person-perceived participation.

Therefore, our objectives were: (1) to determine the strength of the relationship among variables of upper-limb impairment, upper-limb performance in ADL (activity), and person-perceived participation, (2) to determine the upper-limb impairment variables that best explain activity and participation, and (3) to determine the activity variables that best explain participation in people with chronic stroke. We hypothesized that: (1) there would be a significant relationship among variables of upper-limb impairment, activity, and participation, and (2) given that participation can be influenced by a number of factors (eg, environment, culture, motivation), upper-limb impairment variables would explain a larger potion of activity and participation in people with chronic stroke.

Method

Participants

Community-dwelling individuals with chronic stroke and residual unilateral upper-limb impairment were recruited on a voluntary basis using advertisements in community centers and local newspapers. Inclusion criteria included: (1) single stroke of at least 1 year since onset, (2) ability to provide informed consent, and (3) a score of [greater than or equal to] 23 on the Mini-Mental Status Examination. People with significant musculoskeletal conditions (eg, arthritis, previous fracture of the arm that caused deformity, muscle atrophy) or neurological conditions (eg, Parkinson disease, multiple sclerosis, Huntington disease) other than stroke and people with receptive aphasia (ie, based on caregiver information or inability to respond to the instructions "raise your right/left upper limb") were excluded from the study. Ethics approval was obtained from the local university and hospital review boards.

Each participant took part in a 90-minute individual evaluation. An occupational therapist with clinical experience in treating individuals with the residual effects of stroke and one trained research assistant assessed all participants.

A total of 96 participants were screened for the study. Three individuals were excluded after screening due to receptive aphasia; therefore, data for 93 participants were included in the final analysis. The mean age of the participants was 68.7 years (SD=9.4, range=50-93), and 65% were male. More than half (57%) of the participants had left-sided paresis. Further descriptive statistics are presented in Table 1.

Outcome Measures

Measures of impairment. The Modified Ashworth Scale (MAS) (21) was used to measure tone of the paretic elbow flexors. The MAS is an ordinal scale with scores ranging from 0 (normal) to 4 (rigid). The MAS includes a score of 1 + (slight increase in tone with minimal resistance through less than half-range), which is distinctive from a score of 1 where the resistance is felt only at end-range. Interrater reliability has been found to be excellent (intra-class correlation coefficients [ICCs]=.82-.90). (21-23) Validity of MAS scores has been established using isokinetic dynamometry to measure passive resistive force (r=.52-.91). (24,25) The participants were told that the tester would be evaluating arm movement when the arm is relaxed; in this relaxed position, the tester would be moving the arm back and forth at least 5 times. The motion was demonstrated using the tester's arm.

Isometric strength of the paretic upper limb was tested using handheld dynamometry. Wrist and elbow flexion and extension, as well as shoulder flexion and abduction, were assessed. All participants sat in the same chair for the strength testing. Standard upper-limb positions for manual muscle testing were used for the strength testing of each muscle group (eg, for shoulder flexion, shoulder flexed to 90[degrees], elbow flexed to 90[degrees], and shoulder adduction at 0[degrees]). Some participants' joint position would not allow for standard position, so accommodations were made. To ensure standardization of dynamometer placement, tape was placed 5.08 cm (2 in) proximal to the lateral epicondyle, 2.54 cm (1 in) proximal to the medial and radial styloids, and 2.54 cm proximal to the metacarpophalangeal joints on the affected upper limb.

The participants were instructed to push as hard as they could against the dynamometer for 3 seconds and then relax. There was a 10-second rest break between trials. The average of 3 trials for all measures of strength was used to determine the final recorded score. The score from each muscle group tested was summed for a composite score for each participant. We have previously reported isometric strength in people with stroke and found similar average range and magnitude values of these upper-extremity muscles in the paretic limb (26); thus, one muscle group should not have undue weighting on the composite score.

Grip strength of the paretic hand was determined using a Jamar dynamometer. * Each participant sat in the same chair with his or her upper limb in 0 degrees of shoulder adduction and 90 degrees of elbow flexion and wrist between 0 and 30 degrees of flexion. The participant was instructed to squeeze as hard as he or she could for 3 seconds and then relax. The average of 3 trials was used to determine the final recorded score. High interrater reliability (ICC=.88-.99) (26-28) and test-retest reliability (ICC=.80-.98) (29,30) have been found for handheld dynamometry. Validity has been found to be excellent with comparison to known weights (accuracy=1%-3%). (31)

Sensory testing was done using a pressure aesthesiometer kit comprising 8 monofilaments. Sensation was measured on the dorsal lateral aspect of the index finger of the paretic hand. Sensation was tested using filaments presented from thick to fine and deformed to half of their length. The score is established once the individual is not able to detect the pressure of the monofilament. Sham trials (where a filament was not administered, but the participants were asked whether they felt any pressure) were dispersed randomly within each filament presentation. Interrater reliability (ICC=.77-.99) (32) and test-retest reliability (ICC=.69-71) (33,34) for monofilaments has been investigated with satisfactory results.

The Brief Pain Inventory (BPI) (35) was used to assess pain intensity and interference with function (eg, household chores, walking, sleeping). Participants were asked to report whether they had pain of the paretic shoulder, upper limb, and hand only. Each item was rated on an 11-point ordinal scale (0=no pain and 10=worst pain). Internal consistency (Cronbach alpha=.89-.95) (36,37) of the BPI scores was found to be excellent. Validity of the BPI scores assessed with a visual analog scale (r =.66) (36) and the Pain Needs Assessment (r=.60) (37) has been shown to be satisfactory.

Measures of activity. The CAHAI was used to evaluate the capacity of the paretic upper limb in the completion of ADL. The assessor encouraged the participants to use both hands to complete each task. The CAHAI consists of 13 tasks of daily living (eg, pouring, buttoning, zipping). Scoring is done on a 7-point ordinal scale (1=total assistance and 7=complete independence). Scoring is based on the percentage of contribution to each task by the paretic upper limb. For example, an individual would score 7 on the jar-opening task if he or she were able to hold the jar in the nonparetic hand and open it with the paretic hand. A score of 3 means that the individual is able to use the paretic hand to stabilize and manipulate the jar but requires hand-over-hand guidance (50%-74% contribution of the paretic upper limb). High internal consistency (Cronbach alpha=.98) (38) and excellent interrater reliability (ICC=.98), construct validity (r=.81-93), (38) and face and content validity have been reported. (39)

The MAL was used to measure each participant's performance in ADL. The MAL is a semistructured interview that consists of 30 ADL items (eg, brushing teeth, buttoning a shirt, eating). Scoring is completed using 2 scales: (1) Amount of Use scale (0=paretic upper limb is not used and 5=paretic upper limb is used as much as prior to the stroke) and (2) Quality of Movement scale (0=movement quality is poor and 5=movement quality is the same as before the stroke). The MAL has been used as an outcome measure to evaluate upper-limb use by individuals with stroke. (40-41) The MAL also is a useful measure because it evaluates the amount of paretic upper-limb use during ADL, unlike traditional ADL measures in which compensation from the nonparetic upper limb can play a large role in performance. The MAL has been shown to have high internal consistency (Cronbach alpha [greater than or equal to] 88) and reasonable construct validity (Speupper limban's rho=.63) in people with stroke. (42) The MAL has good interrater reliability (ICC=.90-.94). (40,43)

Measure of participation. The RNL Index was used to measure person-perceived participation, (44) which is the individual's perception of his or her involvement in life situations. The RNL Index was developed to measure the effect of disease or trauma on a person's ability to resume ADL, roles, and community functioning. (45) This measure has been widely used as an outcome measure of global functioning, (46) social integration, (47,48) and quality of life (49,50) in individuals with stroke. The RNL Index consists of 11 items, with an emphasis on participation in activities and the community (eg, "I participate in social activities with my family, friends, and/or business acquaintances as is necessary or desirable to me," "I am able to participate in recreational activities as I desire," "I assume a role in my family that meets my needs"). Items are scored on a 3-point ordinal scale (1 =not able to participate as desired and 3=able to fully participate as desired). Good interrater reliability (ICC=.62) and internal consistency (Cronbach alpha= .90-.95) (20) has been found as well as good validity (r=.72) with the Spitzer Quality of Life Index. (45)

Test-retest reliability for the measures used in this study was established using 12 participants with a 1-week interval between tests (ICC= .86-.98). Interrater reliability testing was not done because each tester assessed different outcome measures.

Data Analysis

Descriptive statistics were used to show participant demographics and study measures. Visual inspections of boxplots, histograms, and skewness values were used to determine variable normality and homoscedasticity.

Correlation analysis was used to determine the statistical significance and strength of the relationship among variables of upper-limb impairment, activity, and participation (study objective 1). Correlation analysis was used as the initial step required in determining the variables appropriate for inclusion in the regression analysis. The strength of the relationship between independent and dependent variables is described using the correlation coefficient (r) and was based on Munro's correlation descriptors (51) (very low=.15-24, Iow=.25-.49, moderate=.50-69, high=.70-.89, and very high= .90-1.00). Bivariate correlations were generated using the Pearson product moment correlation for interval data and the Spearman correlation coefficient for ordinal data. Scatterplots of independent variables against dependent variables were visually inspected to determine linearity and to ensure that outlier and influential data points did not compromise the results.

Multiple regression analysis was used to determine which impairment variables best explain upper-limb performance in ADL and participation and which upper-limb performance variables best explain participation in individuals with chronic stroke (study objectives 2 and 3). Those variables that were correlated (ie, related) with upper-limb performance and participation were entered into their respective multiple regression models.

To ensure that the assumptions of multiple regression were met, scatter-plots of residuals against the model data set were inspected, as were tolerance values and the variance inflation factor for possible problems with outliers, influential data points, and multicollinearity. (52) A total of 4 forward stepwise regressions were used to establish models of upper-limb activity and participation. To test the significance of subsets within the regression models, the values of the [R.sup.2] difference test were examined. Variable entry for the regression was set at .05, and removal was set at .10. Four forward regression models were created: 2 models of upper-limb performance (activity) and 2 models of participation (Tab. 2). A value of P[less than or equal to]05 was considered significant in all calculations. The SPSS statistical software, version 11.5 for Windows, ([dagger]) was used for all analyses.

Results

Correlation Results

The Pearson product moment correlations and the Spearman rank correlations are presented in Table 3. Statistically significant correlations were found between upper-limb impairment variables and upper-limb performance variables. Paretic upper-limb strength, MAS score, and grip strength were the impairment variables with the strongest relationship with the CAHAI and MAL scores. The MAS (r=-.23, P<.05), MAL (r=.23, P <.05), and CAHAI (r=.22, P<.05) scores demonstrated weak relationships, but had statistically significant correlations with participation (RNL Index scores).

Regression Results

The first activity model using CAHAI scores as the dependent variable retained the variables of paretic upper-limb strength, MAS scores, sensation, and grip strength ([R.sup.2]=.93, P<.0001) (Tab. 4). In the first step of the model, paretic upper-limb strength accounted for 87% (P< .0001) of the variance of the CAHAI scores. In the second step, the MAS scores contributed an additional 4% to the model, and in the third step grip strength contributed an additional 2%. In the second model of activity with MAL scores as the dependent variable, paretic upper-limb strength was the only variable retained ([R.sup.2]=.78, P < .0001), accounting for 78% of the variance of the MAL scores (Tab. 4).

In the participation model with RNL Index scores as the dependent variable, the MAS scores were the only impairment variable retained ([R.sup.2]=.05, P=.04), accounting for 5% of the variance of the RNL Index scores (Tab. 4). With the RNL Index scores as the dependent variable and the CAHAI scores and MAL scores as the independent variables, the CAHAI scores were the sole variable retained ([R.sup.2]=.05, P=.04), accounting for 5% of the variance of the RNL Index scores.

Discussion

This study examined the strength of the relationship among specific upper-limb impairments and measures of activity (ADL) and participation. Our study sample had a wide range of impairment and activity scores that are typical of community-dwelling individuals with chronic stroke. (7,14,53,54) We were able to determine which impairment variables best explained performance and use of the paretic upper limb using measures exclusive to upper-limb function and not global function. We detected a number of upper-limb impairments that related to activity and participation in individuals with chronic stroke. Namely, we found that paretic upper-limb strength had the strongest relationship with activity, followed by tone and grip strength.

Strength of the paretic upper limb was strongly related to measures of activity and was the strongest contributor in the multivariate models, accounting for 78% to 87% of the variance. This finding illustrates the relationship between weakness and poor performance on measures of ADL; the weaker the paretic upper limb, the worst the score on upper-limb performance measures. Strength involves the capacity to generate sufficient force for movement. Weakness in upper-limb muscles could impair stabilization of proximal arm segments, limit reaching ability, confine hand usage, and affect upper-limb control and coordination. These factors would have a direct effect on the ability and use of the paretic upper limb in daily activities, supporting our findings of the importance of paretic upper-limb strength.

The results of our study and other Studies (12,13,53-55) suggest that the remediation of paretic upper-limb strength may be important for the recovery of ADL independence. Upper-limb strength is cited (12,13,53,54) as a contributing factor to poor upper-limb performance in ADL after stroke; however, few studies have focused on strength training of the paretic upper limb. (55) This highlights the need for randomized control trials to be undertaken to clarify the role of upper-limb strengthening for improvement in upper-limb performance.

Increased tone is a common symptom seen after stroke, and although it is thought to contribute significantly to disability, there is no consensus on whether remediation of increased tone should be a focus of physical therapy or occupational therapy after stroke. (2,56,57) However, promising results for the reduction of tone and improvement in function using pharmacological intervention have been found. (58-60) We found that tone had a strong relationship with activity (upper-limb performance in ADL) and a weak relationship with participation. It also was a statistically significant factor in the models of activity and participation, but accounted for a small proportion of each model (4% in the activity model and 5% in the participation model).

These findings suggest that increased tone does not contribute to poor upper-limb performance in ADL and participation (as measured by our variables), especially compared with paretic upper-limb strength. These findings are not unlike those of other studies (2,56) in which tone, although having a statistically significantly relationship to ADL, had minimal influence when compared with other impairment variables. Although the median of the MAS scores was low in our sample, this finding is typical of studies that have evaluated tone in individuals with chronic stroke. (2,61,62) These studies showed a mean MAS score of less than 1.5. This is also reflective of the low prevalence of high tone scores (>15%) on the MAS (2,61,62) in individuals with chronic stroke.

Paretic grip strength showed a moderate relationship with measures of activity, and although it was statistically significant in the first activity model (CAHAI), it did not contribute substantially (adding 1% to the overall model). Paretic grip strength was not able to explain poor performance in ADL. This finding is in contrast to other studies of individuals with stroke that have shown that grip strength is a significant factor of functional recovery. (12,13,54) In the chronic stage of recovery, individuals may have developed ways to cope with grip strength impairment and thus are able to complete functional tasks.

Boissy et al (12) and Mercier and Bourbonnais (13) evaluated small samples of individuals with chronic stroke (N = 15 and N = 13, respectively) and found grip strength to be a significant contributor to models of upper-limb performance (62%-93% of the variance) in ADL. However, Boissy et al (12) did not include both grip and upper-limb strength in their model, and it is possible that, when grip strength is regressed independently, it is a significant factor in upper-limb performance in ADL, but that, in the presence of other variables (eg, strength, sensation, tone), its effect is minor.

Thirty-seven percent of our study sample reported upper-limb pain, although pain was reported at very low levels (mean of 9.9 out of a possible 120 on the BPI). Shoulder pain after stroke is commonly cited in the literature as both an acute and long-term management issue, with rates reported up to 84%. (63) In a recent population-based study by Ratnasabapathy et al, (64) rates of shoulder pain increased from 17% at 1 week, to 20% at 1 month, to 23% at 6 months after stroke, although additional findings suggest that upper-extremity pain starts to diminish in the chronic stage. (63) Upper-limb pain has been associated with and is considered a prognostic indicator of poor functional recovery. (64,65) One study (63) assessed the relationship of chronic pain after stroke with health-related quality of life, and the authors found no relationship between them.

Because of the lack of definitive findings and the logical sentiment that pain would interfere with ADL performance and engagement in social, recreational, and leisure activities, we chose to include upper-limb pain in our activity and participation models. However, pain was not statistically related to activity or participation. Furthermore, pain was not a variable of any significance in the regression models. The low mean score of our pain measure is consistent with pain scores taken from individuals with chronic stroke. (63,64) These studies suggest that even with the report of pain, it does not interfere significantly with ADL and quality of life. It may be that as the individual recovers, pain diminishes through the natural time course of recovery and the frequency and intensity of reported pain decreases. In addition, individuals in the chronic stage of recovery may adapt to pain levels and thus do not report pain as interfering with performance in ADL or participation.

We found only 2 variables (MAS and CAHAI scores) related to participation. Although studies tend to find a weak relationship (r<.3) between impairment and participation and a moderate relationship (r>.5) between ADL and participation, (66-68) we did not find such a difference. Both impairment and upper-limb activity variables showed a weak relationship and minimal influence on the participation models. The concept of participation is vast, with many elements contributing to its construct. Measuring the contribution of upper-limb impairment and activity may be too limiting. This may indicate that, although upper-limb impairment is relevant to a person's use of his or her upper limb, it is not relevant to his or her involvement in life roles or social activities. It may reflect that individuals in the chronic stage of stroke feel that that they are able to meet their needs given the compensation and adaptation that they have made with their residual impairment and disability.

Limitations

We note several limitations in our study. First, due to the cross-sectional nature of our study, the results can be generalized only to community-dwelling individuals in the chronic stage of stroke recovery. Second, our variable list was not exhaustive, and other variables also may be important (eg, coordination, finger dexterity, force). In our study, we assessed the lateral aspect of the index finger to represent sensory function of the hand; however, this variable may not be reflective of global sensory function of the upper limb. We chose to evaluate this sensory area based on its contribution to hand and pinch motions. Therefore, our results are limited to the finding that sensation assessed from a small area of the hand was not a major contributing factor in our model of upper-limb function.

Our measures of activity focused on a subset of tasks (ADL); therefore, conclusions about upper-limb activity beyond ADL cannot be made. Finally, our participants reported a high level of participation with low variability based on the RNL Index scores. The high scores and low variability of the RNL Index are likely reflective of long-term survivors of stroke rating high satisfaction in their person-perceived participation. (68) However, the low variability of the RNL Index scores may have reduced the strength between the independent variables and the RNL Index scores. Furthermore, the RNL Index measures an individual's perception of his or her involvement in life situations in physical, recreation, and social activities but does not include areas of coping, mood, financial, and environmental supports. Thus, our results are limited to person-perceived satisfaction with physically and socially based activities.

Conclusions

Our hypotheses that there would be a significant relationship among variables of upper-limb impairment, activity, and participation and that upper-limb impairment would explain a larger portion of upper-limb performance compared with participation were confirmed. We found a significant relationship among variables of upper-limb impairment, activity (ADL subset), and person-perceived participation in individuals with chronic stroke. Additionally, paretic upper-limb strength was found to be a strong indicator of upper-limb performance in ADL. However, only weak relationships were found between upper-limb impairment and performance variables with participation. This knowledge can assist clinicians in prioritizing specific treatment interventions (eg, strength training) to enhance upper-limb performance in ADL. The information from this study may provide a basis for initiating randomized controlled trials to evaluate the effectiveness of upper-limb strength training on ADL in individuals with stroke.

* JA Preston Corp, PO Box 89, Jackson, MI 49204.

([dagger]) SPSS Inc, Street233 s Wacker Dr, Chicago, IL 60606.

References

(1) The Growing Burden of Heart Disease and Stroke in Canada, 2003. Ottawa, Ontario, Canada: Heart and Stroke Foundation of Canada; 2003.

(2) Sommerfeld DK, Eek EUB, Svensson A, et al. Spasticity after stroke: its occurrence and association with motor impairments and activity limitations. Stroke. 2004;35: 134-140.

(3) Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Compensation in recovery of upper extremity function after stroke: the Copenhagen Stroke Study. Arch Phys Med Rehabil. 1994;75:852-857.

(4) Richards L, Pohl P. Therapeutic interventions to improve upper extremity recovery and function. Clin Geriatr Med. 1999; 15:819-833.

(5) Wade DT. Measuring upper limb impairment and disability after stroke. Int Disabil Stud. 1989;11:89-92.

(6) Bohannon RW, Andrews AW, Smith MB. Rehabilitation goals of patients with hemiplegia. Int J Rehabil Res. 1988;11:181-183.

(7) Desrosiers J, Malouin F, Bourbonnais D, et al. Upper limb and leg impairments and disabilities after stroke rehabilitation: relation to handicap. Clin Rehabil. 2003;17: 666-673.

(8) Lai SL, Duncan P, Keighley J. Prediction of functional outcome after stroke: comparison of the Orpington Prognostic Scale and the NIH Stroke Scale. Stroke. 1998;29: 1838-1842.

(9) Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Recovery of upper extremity function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil. 1994;75:394-398.

(10) Loewen SC, Anderson BA. Predictors of stroke outcome using objective measurement scales. Stroke. 1990;21:78-81.

(11) Bohannon RW. Motor variables correlated with the hand to mouth maneuver in stroke patients. Arch Phys Med Rehabil. 1991 ;72:682-684.

(12) Boissy P, Bourbonnais D, Carlotti MM, et al. Maximal grip force in chronic stroke subjects and its relationship to global upper extremity function. Clin Rehabil. 1999;13:354-362.

(13) Mercier C, Bourbonnais D. Relative should flexor and handgrip strength is related to upper limb function after stroke. Clin Rehabil. 2004;18:215-221.

(14) Feys J, De Weerdt W, Nuyens G, et al. Predicting motor recovery of the upper limb after stroke rehabilitation: value of a clinical examination. Physiother Res Int. 2000;5:1-18.

(15) International Classification of Functioning, Disability and Health: ICF. Geneva, Switzerland: World Health Organization; 2001.

(16) Barreca SR, Gowland C, Stratford PW, et al. Development of the Chedoke Upper Limb and Hand Activity Inventory: item selection. Physiother Can. Summer 1999: 209-211.

(17) Taub E, Miller NE, Novack TA, et al. Technique to improve chronic motor deficit after stroke. Arch Phys Med Rehabil. 1993;74: 347-354.

(18) International Classification of Impairments, Disabilities, and Handicaps: A Manual of Classification Relating to the Consequences of Disease. Geneva, Switzerland: World Health Organization; 1980.

(19) Carr AJ, Thompson PW. Towards a measure of patient-perceived handicap in rheumatoid arthritis. Br J Rheumatol. 1994;33:378-382.

(20) Wood-Dauphinee SL, Williams JI. Reintegration to normal living index as a proxy to quality of life. J Chronic Dis. 1987;40: 491-499.

(21) Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther. 1987;67:206-207.

(22) Allison SC, Abraham LD, Peterson CL. Reliability of the Modified Ashworth Scale in the assessment of flexor muscle spasticity in patients with traumatic brain injury. Int J Rehabil Res. 1996;19:67-78.

(23) Sloan RL, Sinclair E, Thompson J, et al. Interrater reliability of the Modified Ashworth Scale for spasticity in hemiplegic patients. Int J Rehabil Res. 1992;15:158-161.

(24) Pandyan AD, Price CIM, Barnes MP, Johnson GR. A biomechanical investigation into the validity of the Modified Ashworth Scale as a measure of elbow spasticity. Clin Rehabil. 2003;17:290-294.

(25) McCrea PH, Eng JJ, Hodgson AJ. Linear spring-damper model of the hypertonic elbow: reliability and validity. J Neurosci Methods. 2003;128:121-128.

(26) McCrea PH, Eng JJ, Hodgson AJ. Time and magnitude of torque generation is impaired in both upper limbs following stroke. Muscle Nerve. 2003;28:46-53.

(27) Bohannon RW. Internal consistency of dynamometer measurements in healthy subjects and stroke patients. Percept Mot Skills. 1995;81:113-114.

(28) Bohannon RW. The clinical measurement of strength. Clin Rehabil. 1987;1:5-16.

(29) Mathiowetz V. Comparison of Rolyan and Jamar dynamometers for measuring grip strength. Occup Ther Int. 2002;9:201-209.

(30) Ottenbacher KJ, Branch LG, Ray L, et al. The reliability of upper- and lower-extremity strength testing in a community survey of older adults. Arch Phys Med Rehabil. 2002 ;83:1423-1427.

(31) LaStayo P, Hartzel J. Dynamic versus static grip strength: how grip strength changes when the wrist is moved, and why dynamic grip strength may be a more functional measurement. J Hand Ther. 1999;12:212-218.

(32) Novak CB, Mackinnon SE, Williams JI, Kelly L. Establishment of reliability in the evaluation of hand sensibility. Plast Reconstr Surg. 1993;92:311-322.

(33) Halar EM, Hammond MC, LaCava EC, et al. Sensory perception threshold measurement: an evaluation of semiobjective testing devices. Arch Phys Med Rehabil. 1987;68:499-507.

(34) Sieg KW, Williams WN. Preliminary report of a methodology for determining tactile location in adults. Occup Ther J Res. 1986; 6:195-206.

(35) Cleeland CS, Ryan KM. Pain assessment: global use of the Brief Pain inventory. Ann Acad Med. 1994;23:129-138.

(36) Tyler EJ, Jensen MP, Engel JM, Schwartz L. The reliability and validity of pain interference measures in persons with cerebral palsy. Arch Phys Med Rehabil. 2002;83: 236-239.

(37) Wang XS, Mendoza TR, Goa SZ, Cleedland CS. The Chinese version of the Brief Pain Inventory (BPI-C): its development and use in a study of cancer pain. Pain. 1996; 67:407-416.

(38) Barreca SR, Stratford PW, Lambert CL, et al. Test-retest reliability, validity, and sensitivity of the Chedoke Upper Limb and Hand Activity inventory: a new measure of upper limb function for survivors of stroke. Arch Phys Med Rehabil. 2005; 86:1616-1622.

(39) Barreca SR, Gowland C, Stratford PW, et al. Development of the Chedoke Upper Limb and Hand Activity Inventory: theoretical constructs, item generation, and selection. Top Stroke Rehabil. 2004;11: 31-42.

(40) Miltner W, Bauder H, Sommer M, et al. Effects of constraint-induced movement therapy on patients with chronic motor deficits after stroke. Stroke. 1999;30: 586-592.

(41) Liepert J, Baude H, Miltner W, et al. Treatment-induced cortical reorganization after stroke in humans. Stroke. 2000;31: 1210-1216.

(42) Van der Lee J, Beckerman H, Knol DL, et al. Clinimetric properties of the Motor Activity Log for the assessment of upper limb use in hemiparetic patients. Stroke. 2004;35:1410-1414.

(43) Taub E, Uswatte G, Pidikitti RP. Constraint-induced movement therapy: a new family of techniques with broad application to physical rehabilitation--a clinical review. J Rehabil Res Dev. 1999;36: 237-251.

(44) Cardol M, Brandsma JW, de Goot IJ, et al. Handicap questionnaires: what do they assess? Disabil Rehabil. 1999;21:97-105.

(45) Wood-Dauphinee SL, Opzoomer A, Williams JI, et al. Assessment of global function: the Reintegration to Normal Living Index. Arch Phys Med Rehabil. 1988;69: 583-590.

(46) Chen Y, Rodger S, Polatajko H. Experiences with the COPM and client-centred practice in adult neurorehabilitation in Taiwan. Occup Ther lnt. 2002;9:167-184.

(47) Chung RY, Carter BS, Norbash A, et al. Management outcomes for ruptured and unruptured aneurysms In the elderly. Neurosurgery. 2000;47:827-833.

(48) Clarke PJ, Black SE, Badley EM, et al. Handicap in stroke survivors. Disabil Rehabil. 1999;21:116-123.

(49) Bethoux F, Calmels P, Gautheron V. Changes in the quality of life of hemiplegic stroke patients with time: a preliminary report. Am J Phys Med Rehabil. 1999;78: 19-23.

(50) Bluvol A, Ford-Gilboe M. Hope, health, work and quality of life in families of stroke survivors. J Adv Nurs. 2004;48:322-332.

(51) Munro BH. Correlations. In: Munro BH, Visintainer MA, Page EB, eds. Statistical Methods for Health Care Research. Philadelphia, Pa: JB Lippincott Co; 1993:181.

(52) Cohen J, Cohen P, West SG, Aiken LS. Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. 3rd ed. Mahwah, NJ: Lawrence Erlbaum Associates; 2003.

(53) Bourbonnais D, Bilodeau S, Lepage Y, et al. Effect of force-feedback treatments in patients with chronic motor deficits after stroke. Am J Phys Med Rehabil. 2002 ;81:890-897.

(54) Sunderland A, Tinson D, Bradley L, Hewer R. Upper limb function after stroke: an evaluation of grip strength as a measure of recovery and a prognostic indicator. J Neurol Neurosurg Psychiatry. 1989;52: 1267-1272.

(55) Eng JJ. Strength training in individuals with stroke. Physiother Can. 2004;56: 1-12.

(56) O'Dwyer NJ, Ada L, Neilson PD. Spasticity and muscle contracture following stroke. Brain. 1996;119:1737-1749.

(57) Bobath B. Adult Hemiplegia." Evaluation and Treatment. Oxford, United Kingdom: Butterworth-Heinemann; 1990.

(58) Woldag H, Hummelsheim H. Is the reduction of spasticity by botulinum toxin A beneficial for the recovery of motor function of upper limb and hand in stroke patients? Eur Neurol. 2003;50: 165-171.

(59) Miscio G, Del Cotne C, Pianca D, et al. Botulinum toxin in post-stroke patients: stiffness modifications and clinical implications. J Neurol. 2004;251:189-196.

(60) Thompson AJ, Jarrett L, Lockley L, et al. Clinical management of spasticity. J Neurol Neurosurg Psychiatry. 2005;76:459-463.

(61) Watkins CL, Leathley MJ, Moore AP, et al. Prevalence of spasticity post stroke. Clin Rehabil. 2002;16:515-522.

(62) Pang YC, Harris JE, Eng JJ. A community-based upper extremity exercise program improves motor control and performance of functional activities in chronic stroke: a pilot randomized controlled study. Arch Phys Med Rehabil. 2006;87: 1-9.

(63) Kong KH, Woon VC, Yang SY. Prevalence of chronic pain and its impact on health-related quality of life in stroke survivors. Arch Phys Med Rehabil. 2004;85:35-40.

(64) Ratnasabapathy Y, Broad J, Baskett J, et al. Shoulder pain in people with stroke: a population-based study. Clin Rehabil. 2003;17:304-311.

(65) Roy CW, Sands MR, Hill LD. Shoulder pain in acutely admitted hemiplegics. Clin Rehabil. 1994;8:334-340.

(66) De Haan R, Horn J, Limburg M, et al. A comparison of five stroke scales with measures of disability, handicap, and quality of life. Stroke. 1993;24:1178-1181.

(67) Gottlieb A, Golander H, Bar-Tal Y, Gottlieb D. The influences of social support and perceived control on handicap and quality of life after stroke. Aging Clin Exp Res. 2001;13:11-15.

(68) Hackett ML, Duncan JR, Anderson CS, et al. Health-related quality of life among long-term survivors of stroke: results from the Auckland Stroke Study, 1991-1992. Stroke. 2000;31:440-447.

JE Harris, OT, MSc, is a graduate student in the School of Rehabilitation Sciences, University of British Columbia, Vancouver, British Columbia, Canada, and in the Rehabilitation Research Lab, G.F. Strong Rehab Centre, Vancouver, British Columbia, Canada.

JJ Eng, PT/OT, PhD, is Professor, School of Rehabilitation Sciences, University of British Columbia, T325-2211 Wesbrook Mall, Vancouver, British Columbia, Canada V6T 2B5, and Scientist, Rehabilitation Research Lab, G.F. Strong Rehab Centre. Address all correspondence to Dr Eng at: Janice. [e-mail].

[Harris JE, Eng JJ. Paretic upper-limb strengthbest explains arm activity in people with stroke. Phys Ther. 2007;87:88-97.]

[C] 2007 American Physical Therapy Association

Ms Harris provided study concept and design, data collection and analysis, and manuscript preparation. Dr Eng provided project management, development of study design and methodology, assistance with data analysis, and preparation of the manuscript. The authors acknowledge Dr Joanne Desrosiers for her review of the manuscript and her useful comments and suggestions.

This work was presented at the World Stroke Congress; June 24, 2004; Vancouver, British Columbia, Canada.

This work was supported by an operating grant from the Heart and Stroke Foundation of British Columbia and Yukon and career scientist awards to Dr Eng from the Canadian Institute of Health Research (CIHR) and the Michael Smith Foundation for Health Research and by a CIHR Doctoral Award and a Strategic Training Fellowship in Rehabilitation Research from the CIHR Musculoskeletal and Arthritis Institute to Ms Harris.

This article was received February 28, 2006, and was accepted August 23, 2006.

DOI: 10.2522/ptj.20060065

Table 1.
Descriptive Statistics

Variable                                         n            [bar.x]

Time since stroke (y)                                           5.1
Stroke type (ischemic/hemorrhagic/            34/18/41
  unknown)
Mini-Mental Status                                             26.0
  Examination (0-30)
Fugl-Meyer Upper Extremity                                     44.0
  Motor Scale (0-66)
Modified Ashworth Scale,                                        0.90
  elbow (0-4)
Upper-limb strength (kg)                                       45.4
Grip strength (kg)                                             13.0
Sensation (1-8)                                                 4.1
Brief Pain Inventory                                            9.9
  total score (0-120)
Chedoke Arm and Hand                                           62.1
  Activity Inventory (13-91)
Motor Activity Log                                              3.1
Reintegration to Normal                                        29.2
  Living Index (11-33)

Variable                                       SD            Range

Time since stroke (y)                         4.1              1-27
Stroke type (ischemic/hemorrhagic/
  unknown)
Mini-Mental Status                            3.0             23-29
  Examination (0-30)                         21.1              4-66
Fugl-Meyer Upper Extremity
  Motor Scale (0-66)                          1.0 (median)     0-4
Modified Ashworth Scale,                     29.4            0.0-132.5
  elbow (0-4)                                11.1            0.0-43.7
Upper-limb strength (kg)                      2.2              1-8
Grip strength (kg)                           17.3              0-88
Sensation (1-8)                              31.8             13-91
Brief Pain Inventory
  total score (0-120)                         1.6              0-5
Chedoke Arm and Hand                          3.6             19-33
  Activity Inventory (13-91)
Motor Activity Log
Reintegration to Normal
  Living Index (11-33)

Table 2.
Regression Models: Two Activity and Two Participation Models

                             Activity
                             Model 1

Dependent variable           Chedoke Arm and Hand
                               Activity Inventory
Independent variables        Impairment measures:
                               Upper-limb strength
                               Modified Ashworth Scale
                               Grip strength
                               Sensation
                               Brief Pain Inventory

                             Activity
                             Model 2

Dependent variable           Motor Activity Log

Independent variables        Impairment measures:
                               Upper-limb strength
                               Modified Ashworth Scale
                               Grip strength
                               Sensation
                               Brief Pain Inventory

                             Participation
                             Model 1

Dependent variable           Reintegration to Normal
                               Living Index
Independent variables        Impairment measures:
                               Upper-limb strength
                               Modified Ashworth Scale
                               Grip strength
                               Sensation
                               Brief Pain Inventory

                             Participation
                             Model 2

Dependent variable           Reintegration to
                               Normal Living Index
Independent variables        Activity measures:
                               Chedoke Arm and
                                 Hand Activity
                                 Inventory
                             Motor Activity Log

Table 3.
Correlations Among Upper-Limb Impairment, Activity, and Participation
Variables (a)

Variable                        Chedoke Arm and         Motor
                                 Hand Activity       Activity Log
                                   Inventory

Modified Ashworth                   -.80 **             -.71 **
  Scale (b)
Upper-limb strength (c)              .89 **              .84 **
Grip strength (c)                    .69 **              .61 **
Sensation (b)                       -.42 **             -.43 **
Brief Pain Inventory (c)            -.03                -.06
Chedoke Arm and                                          .82 **
  Hand Activity
  Inventory (c)
Motor Activity Log (b)

Variable                       Reintegration to
                                 Normal Living
                                     Index

Modified Ashworth                   -.23 *
  Scale (b)
Upper-limb strength (c)              .19 *
Grip strength (c)                    .10
Sensation (b)                       -.09
Brief Pain Inventory (c)            -.04
Chedoke Arm and                      .22 *
  Hand Activity
  Inventory (c)
Motor Activity Log (b)               .23 *

(a) Asterisk indicates P <.05, double asterisk indicates P <.01.

(b) Spearman rank correlation.

(c) Pearson product moment correlation.

Table 4.
Regression Models of Upper-Limb Activity and Participation Using
Impairment Variables

Variable                        [R.sup.2] (a)      [R.sup.2]       P
                                                   Change (b)

Model 1-activity               Chedoke Arm and Hand Activity Inventory
Upper-limb strength                  .87              .87        .0001
Modified Ashworth Scale              .91              .05        .0001
Grip strength                        .93              .004       .03
Model 2-activity                              Motor Activity Log
Upper-limb strength                  .78              .78        .0001
Model 1-participation           Reintegration to Normal Living Index
Modified Ashworth Scale              .05              .05        .04

(a) [R.sup.2] = the proportion of variability of the dependent variable
that was explained by the independent variable.

(b) [R.sup.2] change = how much additional variabilty was explained by
each independent variable.