Fugl-Meyer Assessment of Sensorimotor Recovery After Stroke (FMA)

Overview

The FMA has been used as the gold standard against which the validity of other measures has been assessed. However, the reliability and validity of the Balance subscore (the sitting balance items in particular) of the FMA has been shown to be questionable. As mentioned in the available versions section, revisions to the scoring of the Balance subscore appear to have resulted in an increase in reliability (Mao et al. 2002), however, further testing of the modification is required. The Sensation subscore of the FMA has also been criticized for poor face, construct and predictive validity and responsiveness (Lin, Hsueh, Sheu, & Hsieh, 2004).

Floor/Ceiling Effects

Lin et al. (2004) examined the psychometric properties of the FMA Sensation subscore and found that these subscore had large ceiling effects at each assessment time. At 14-30 days post-stroke, 44.4% of the patients achieved the highest score, at 30-90 days, 48.9%, at 90-180 days, 62.7% and at 14-180 days, 72.1%.

Mao et al. (2002) compared the psychometric properties of the Berg Balance Scale, the modified Balance subscore of the FMA, and the Postural Assessment Scale for Stroke Patients in 123 stroke patients followed up prospectively 14, 30, 90, and 180 days after stroke onset. The modified Balance subscore of the FMA showed large floor effects (29.3%) at 14 days after stroke.

Hsueh, Hsu, Sheu, Lee, Hsieh and Lin (2009) analyzed the floor and ceiling effects for the FMA, the shortened version of the FMA, the Stroke Rehabilitation Assessment of Movement (STREAM), and the shortened version of the STREAM in 50 clients with chronic stroke. Participants were assessed at admission and discharge from a rehabilitation ward. At admission, the FMA and the shortened version of the STREAM demonstrated an excellent floor effect and ceiling effect with no participants scoring the minimum or maximum scores. The other measures showed adequat floor and ceiling effects with 2 to 18% of patients scoring the lowest or highest scores. At discharge, the FMA and the shortened version of the STREAM demonstrated an excellent floor effect with 0% of participants scoring 0. The other measures showed adequat floor and ceiling effects, with the proportion of patients scoring the minimum and maximum scores ranging from 1 to 20%.

Reliability

Lin et al. (2004) examined the internal consistency of the FMA in 176 patients with stroke from 14 to 180 days after stroke. Cronbach’s alphas for the FMA at four time points post-stroke were excellent, ranging from alpha = 0.94 to 0.98. The inter-rater reliability of the total score of the FMA was also excellent, with an intraclass correlation coefficient (ICC) of 0.93. However, item-level agreement for light touch items on the FMA Sensation subscale ranged from poor to adequate (weighted kappa ranged from 0.30 to 0.55).

Platz, Pinkowski, van Wijck, Kim, di Bella, and Johnson (2005) tested the test-retest and the inter-rater reliabilities of the FMA upper extremity items (including items from the Motor function, Sensation and passive Joint motion/Joint pain subscores), the Action Research Arm Test, and the Box and Block Test in patients with upper limb paresis either from stroke (n=37), multiple sclerosis (n=14), or from traumatic brain injury (n=5). Test-retest reliability of the FMA, calculated using ICC’s, was excellent (ICC = 0.97 for Total motor score; ICC = 0.81 for Sensation, ICC = 0.95 for passive Joint motion/Joint pain). Inter-rater reliability for the FMA upper extremity subscore as calculated using the ICC was also excellent (ICC = 0.99 for Total motor score; ICC = 0.98 for Sensation; and ICC = 0.98 for passive Joint motion/Joint pain).

Mao et al. (2002) compared the psychometric properties of the Berg Balance Scale, the modified Balance subscore of the FMA, and the Postural Assessment Scale for Stroke Patients, in 123 patients with stroke followed up prospectively 14, 30, 90, and 180 days after stroke onset. The median of weighted kappa statistics for each item of the FMA Balance subscore was 0.79 (ranging from 0.71 to 0.95), indicating excellent individual item agreement. The ICC for the total score of the FMA Balance subscore was 0.92 (ranging from 0.88 to 0.95), indicating excellent total score agreement. The Cronbach’s alpha for the FMA Balance subscore ranged from alpha = 0.85 to 0.91 on all follow-up times, indicating high internal consistency.

Duncan, Propst, and Nelson (1983) examined the test-retest reliability and the inter-rater reliability of the FMA in 18 patients with chronic stroke. Inter-rater reliability was examined with 5 different therapists. Pearson correlations between therapists for each component of the FMA Motor domain upper extremity subscale were found to be excellent, ranging from r = 0.96 to r = 0.97. The Motor domain lower extremity subscale correlations were also excellent, ranging from r = 0.83 to r = 0.95. The Total score correlation was high (r = 0.99). Only the reflex and coordination subscores in the upper extremity were found to be unreliable as they were significantly different between the raters across the test times.

To assess the test-retest reliability of the FMA in Duncan et al. (1983), one therapist evaluated patients on three separate occasions at 3-week intervals. Pearson correlations were excellent for the Total FMA score (r = 0.98 to r = 0.99), Motor domain upper extremity subscore (r = 0.995 to r = 0.996), Motor domain lower extremity subscore (r = 0.96), Sensation subscore (r = 0.95 to r = 0.96), Joint range of motion/Joint pain subscore (r = 0.86 to r = 0.99) and Balance subscore (r = 0.89 to r = 0.98). No significant differences across evaluation times were found (as assessed by a repeated measures analysis of variance).

Sanford, Moreland, Swanson, Stratford, and Gowland (1993) examined the inter-rater reliability of the FMA in 12 patients between 6 days to 6 months post-stroke. Patients were evaluated one day apart by three physical therapists. The inter-rater reliability of the FMA was found to be excellent, with an overall ICC of 0.96 for the Total score. The ICCs for the Motor domain upper extremity subscore, Motor domain lower extremity subscore, Balance, Sensation, and Joint range of motion were excellent (0.97, 0.92, 0.93, 0.85 and 0.85, respectively). The ICC for the Joint pain subscore was the least reliable, but still adequate, with an ICC of 0.61.

Beckerman, Vogelaar, Lankhorst, and Verbeek (1996) examined the test-retest reliability of the FMA in 49 patients with chronic stroke. Patients were evaluated twice by one therapist, three weeks apart. The ICC for the Motor domain lower extremity subscore was excellent (ICC = 0.86), but was found to be poor for the Balance subscore (ICC = 0.34). Although the test-retest reliability of the Balance subscore was not found to be very reliable, indicating that the patients’ performance was inconsistent, the inter-rater agreement was high.

Van der Lee, Beckerman, Lankhorst and Bouter (2001) examined the test-retest reliability of the FMA Motor domain upper extremity subscore in 22 patients with chronic stroke. Two baseline measurements were performed before treatment, as well as a follow-up measurement after 2 weeks. Using limits of agreement as a measure of test-retest reliability, a mean difference on test-retest within a stable population during the baseline period (14-20 days) on the Motor domain upper extremity subscore was small (0.8 points).

Woodbury, Velozo, Richards, Duncan, Studenski, and Lai (2008) investigated the test-retest reliability of the FMA through longitudinal stability of the upper extremity items in 377 clients with stroke with a mean age of 60.2 years (SD 11.2). Longitudinal stability reflects the invariance of an item-difficulty hierarchy of a measure across 2 testing occasions. Assessments with a slightly modified FMA (the 3 reflex items were removed with the remaining 30 items arranged in a hierarchical sequence) were performed at admission and at 6 months post-stroke. Longitudinal equivalence of the FMA item structure was analyzed by estimating the relationship between item-difficulty calibrations at baseline and 6 months follow-up. Intraclass Correlation Coefficient (ICC) was performed and showed excellent reliability (ICC = 0.95). Consistency of item-difficulty calibrations across testing occasions disregarding clients’ characteristics such as age and gender, as calculated using Differential Item Functioning (DIF), showed that only two items on the FMA (1: shoulder flexion to 180° with elbow extended and 2: movement of the arm with normal speed) had a large DIF, which implied that those two items were unstable between assessments. Despite those items, the authors reported that the scores across testing occasions could still be compared as a repeat Rasch analysis with the unstable items removed found a small difference on the FMA scores (0.06 logits).

Hsueh et al. (2009) analyzed the test-retest reliability of the FMA, the STREAM, and their shortened versions in 60 clients with chronic stroke. Participants were assessed twice within a 1 week interval. Test-retest reliability, as calculated using Intraclass Correlation Coefficient was excellent for all four measures: FMA (ICC = 0.98), the shortened version of the FMA (ICC = 0.96), the STREAM (ICC = 0.98) and the shortened version of the STREAM (ICC = 0.97).

Sullivan et al. (2011) developed a standardized measurement method and rater training program for the FMA and then used inter-rater and intra-rater reliability to examine the effectiveness of the program. After attending the training program, 17 physiotherapists and one expert rater (with 30 years experience) evaluated 15 patients with subacute stroke. The initial assessments were video-recorded and later reviewed by the expert rater. The intra-rater reliability of the expert rater, as calculated using ICC was excellent for all domains of the FMA (ranging from ICC = 0.95 to 1.0). The inter-rater reliability between the expert rater and trained therapists, as calculated using ICC, was also excellent for all domains of the FMA (ranging from ICC = 0.87 to 0.99).

Validity

Content:
Woodbury, Velozo, Richards, Duncan, Studenski, and Lai (2008) investigated the content validity of the upper extremity items of the FMA in 377 clients with stroke with a mean age of 60 years (SD 11,2). Assessments with a slightly modified FMA (the 3 reflex items were removed with the remaining 30 items arranged in a hierarchical sequence) were performed at admission and at 6 months post-stroke. At the follow-up evaluation, the 30-item Fugl-Meyer Assessment, as calculated using Rasch Analysis, showed an acceptable fit statistics, except for the item hook grasp. This result suggests that all items reflect the same construct, except for the item hook grasp.

Crow & Harmeling-van der Wel (2008) analyzed the content validity of the motor functioning domain of the FMA in 62 clients with stroke by performing a Guttmann Scale Analysis. Each motor functioning subscale of the FMA, excluding the reflex items, was arranged in a hierarchical sequence of difficulty according to the total number of passes per item. All subscales, when analyzed separately, exceed the critical values for two indices: coefficient of reproducibility (> 0,9) and coefficient of scalability (>0,7). When analyzing across all upper and lower extremity items (ignoring the subscales) the coefficient of reproducibility (0,8) and scalability (0,6) were just below the acceptable levels. The results of this study suggest the existence of a valid, cumulative, and unidimensional Guttman scale within each motor functioning subscale. In summary, if the client succeeds in completing the most difficult item in a subscale, this suggests he/she will succeed in the easier items for that same subscale. Similarly, failure on an item suggests the client will be unable to complete the remaining more challenging items in the subscale. Therefore, a shortened method of administration of the FMA can be used with clients with stroke.

Criterion:
There is a lack of information available regarding the criterion validity of the FMA, as the FMA was created at a time when no other similar impairment index existed against which the FMA could be tested (Gowland, Van Hullenaar, & Torresin, 1995; Gowland, Stratford, & Ward, 1993).

Bernspang et al. (1987) used the Pearson correlation coefficient and found that the FMA correlated excellently with self-care ability scores (r = 0.64) in 109 clients measured within 2 weeks of stroke.

Nadeau et al. (1999) identified the most important clinical variables for determining gait speed in 16 patients with chronic stroke using Pearson’s correlation coefficients. The FMA Sensation subscore correlated poorly with both comfortable and maximal gait speeds (r = 0.14 and r = 0.05, respectively). Patients with decreased sensation and a score below 12, combined with the strength of the hip flexors, and ankle plantar flexors predicted maximal gait speed. Excellent correlations were found between the FMA Total motor score and comfortable (mean 0.76 meters/second; r = 0.61) and maximal speeds (mean 1.09 meter/second; r = 0.61).

Concurrent:
Wood-Dauphinee et al. (1990) compared the FMA to the Barthel Index in 167 patients with stroke at two time points: the acute stage (3 to 5 days post-stroke), and 5 weeks post-stroke. Using Pearson correlation coefficients, the correlation between the FMA Motor domain upper extremity subscore and the Barthel Index total score was excellent at both the acute stage (r = 0.75) and at 5 weeks (r = 0.82). Similarly, excellent correlations were found between the FMA Motor domain lower extremity subscore and the Barthel Index at the acute stage (r = 0.77) and at 5 weeks (r = 0.89).

Poole and Whitney (1988) examined the concurrent validity of the FMA with the Motor Assessment Scale in 30 patients with stroke. High correlations were found between the total scores on the FMA and the Motor Assessment Scale (r = 0.88), and between specific item scores, except sitting balance (ranging from r = 0.28 to r = 0.92).

Malouin et al. (1994) administered the FMA and the Motor Assessment Scale to 32 patients early after stroke, and reported an excellent Spearman correlation for Total FMA and Total Motor Assessment Scale scores (r = 0.96). The correlations for items from the Motor Assessment Scale and corresponding FMA items were excellent, ranging from r = 0.65 to r = 0.93. The FMA Sensation scores of light touch (r = 0.64) and position sense (r = 0.67) correlated with the Motor Assessment Scale’s Balance score, but not with FMA Sitting balance items (r = 0.12 and -0.10 respectively) suggesting that the FMA sitting balance test is not valid for measuring balance. Standing balance correlations ranged from moderately to highly correlated with both the Motor Assessment Scale and FMA light touch and position sense (ranging from r = 0.43 to r = 0.67).

Di Fabio and Badke (1990) examined standing balance and dynamic weight shifting in 10 patients with stroke using a sensory organization balance test and the FMA. The results of both clinical tests were compared to determine whether the sensory organization balance test correlated with functional ability using the Spearman rank order correlation coefficient. Scores on the lower extremity items and balance items of the FMA correlated excellently with the Sensory Organization Balance Test (r = 0.77).

De Weerdt and Harrison (1985) compared the Motor domain upper extremity subscore of the FMA with the Action Research Arm Test. Both assessments were administered to 53 hospital inpatients with stroke who suffered a motor deficit. The Action Research Arm Test and the FMA were excellently correlated at both 2 weeks (r = 0.91) and at 8 weeks (r = 0.94) post-stroke.

Berglund and Fugl-Meyer (1986) compared the FMA to the DeSouza scale (another assessment of upper limb function) in 50 patients with stroke who suffered a motor deficit. Excellent correlations were found with the Upper extremity Motor scores (r = 0.90). The two tests co-varied, and explained 90% of the variation in the Total scores and 80% of the Motor scores.

Gowland et al. (1993) demonstrated the concurrent validity of the Chedoke-McMaster Stroke Assessment scale with the FMA. The Total score of the impairment inventory of the Chedoke-McMaster Stroke Assessment Scale correlated highly with that of the FMA (r = 0.95). Correlations reported between impairment subscores on the FMA and corresponding subscores from the impairment inventory of the Chedoke-McMaster Stroke Assessment ranged from adequate to excellent (r = 0.76 to r = 0.95).

Mao et al. (2002) compared the Berg Balance Scale, the modified Balance subscore of the FMA, and the Postural Assessment Scale for Stroke, in 123 patients with stroke followed up prospectively at 14, 30, 90, and 180 days after stroke onset. There was excellent concurrent validity (as measured by Spearman correlation coefficient) between the Balance subscore of the FMA and the Berg Balance Scale and Postural Assessment Scale for Stroke at all follow-up times (ranging from r = 0.90 to r = 0.97).

Kusoffsky, Wadell and Nilsson (1982) reported a relationship between sensory functioning and subsequent motor recovery, as excellent correlations between sensory evoked potentials in 16 patients with stroke was observed. The relationship between sensory evoked potentials and the FMA upper extremity subscore was the most strong, and the relationship between the FMA lower extremity subscore was less strong. The strength of the relationship endured regardless of when the FMA was administered.

Feys, Van Hees, Bruyninck, Mercelis and De Weerdt (2000) assessed the role of sensory evoked potentials and motor evoked potentials in the prediction of arm motor recovery in 64 patients with stroke with a motor deficit of the arm. Patients were followed from 2 weeks to 12 months post-stroke. In this study, a poor relationship was found between sensory evoked potentials and the FMA.

Dettmann et al. (1987) administered the FMA and a walking performance test using interrupted light photography and postural maneuvers while standing on a force platform in 15 patients with stroke. Correlations between the FMA Total scores and performance assessments of walking velocity (r = 0.64), cadence (r = 0.58), stride-length (r = 0.53), non-paretic stance (r = 0.59), velocity index (r = 0.67), and platform measures of upright stability (r = 0.52 to r = 0.69) ranged from adequate to excellent.

Poole and Whitney (1988) examined the concurrent validity of the Motor Assessment Scale and the FMA in 30 patients with stroke. Excellent correlations between FMA items and corresponding MAS items were found (ranging from r = 0.64 to r = 0.92). However, one exception was the correlation of the FMA sitting balance with the MAS balance while sitting, which were poorly correlated (r = 0.28).

Chae, Labatia, and Yang (2003) evaluated the concurrent validity of the Arm Motor Ability Test using the FMA as the criterion measure of post-stroke upper limb motor impairment in 30 patients with chronic stroke. Excellent Spearman correlations between upper extremity scores of the FMA and Arm Motor Ability Test functional ability scores (r = 0.94) as well as between upper extremity scores of the FMA and Arm Motor Ability Test quality of movement scores (r = 0.94) were reported.

Hsueh et al. (2009) analyzed the concurrent validity of the FMA, the shortened version of the FMA, the STREAM, and the shortened version of the STREAM in 50 clients with chronic stroke. Excellent Spearman correlations were found between all four measures (ranging from rho = 0.91 to rho = 0.99).

Predictive:
Mao et al. (2002) compared the psychometric properties of the Berg Balance Scale, the modified Balance subscore of the FMA, and the Postural Assessment Scale for Stroke in 123 patients with stroke at 14, 30, 90, and 180 days after stroke onset. Excellent Spearman correlations were found between the scores of the Balance subscore of the FMA at the earlier 3 time points and the Postural Assessment Scale for Stroke scores (ranging from r = 0.80 to r = 0.87), indicating excellent predictive validityof the FMA modified Balance subscore.

Chae, Johnston, Kim, and Zorowitz (1995) found that FMA lower extremity admission subscores predicted the rehabilitation discharge Functional Independence Measure mobility (r = 0.63) and locomotion (r = 0.74) scores in 48 patients at 6 weeks post-stroke.

Hsueh et al. (2009) analyzed whether the motor scores of FMA, of the shortened version of the FMA, the STREAM, and the shortened version of the STREAM measured at admission to a rehabilitation program were able to predict Barthel Index scores at discharge in 50 clients with chronic stroke. Excellent Spearman correlations were found between the Barthel Index scores and the FMA (rho = 0.72), the shortened version of the FMA (rho = 0.74), the STREAM (rho = 0.75) and the shortened version of the STREAM (rho = 0.77). These results suggest that all measures were able to predict Barthel Index scores at discharge.

Fulk, Reynolds, Mondal & Deutsch (2010) examined the predictive validity of the 6MWT and other widely used clinical measures (FMA-LE, self-selected gait-speed, SIS and BBS) in 19 patients with stroke. The FMA-LE was found to be a poor predictor of mean steps per day (r = 0.06; p = 0.798). Although gait speed and balance were related to walking activity, only the 6MWT was found to be a predictor of community ambulation in patients with stroke.

Construct:
Several validation studies have provided good evidence that the FMA is measuring what it is intended to measure. The construct validity of the FMA has been examined by comparing the scale with other measures of stroke recovery that reflect post-stroke independence in activities of daily living or disability level.

Sonde, Gip, Fernaeus, Nilsson, and Viitanen (1998) found that FMA scores reflected changes in upper extremity scores after low frequency transcutaneous electric nerve stimulation. A Mann-Whitney U-test was used to test the significance of the differences in FMA scores at the start and end of the study between the treatment and control groups. This test revealed that the FMA scores differed significantly between the treatment and control groups (Mann-Whitney U = 130.5). Further, the Spearman rank correlation between the FMA scores was excellent (r = 0.95).

Chae, Bethoux, Bohine, Dobos, Davis, and Friedl (1998) assessed the efficacy of neuromuscular stimulation in enhancing the upper extremity motor and functional recovery of 28 patients with stroke. FMA scores reflected changes in upper extremity scores after neuromuscular stimulation. Parametric analyses revealed significantly greater gains in FMA scores for the treatment group immediately following treatment (13.1 versus 6.5), at 4 weeks after treatment (17.9 versus 9.7), and at 12 weeks after treatment (20.6 versus 11.2). Gains were observed in FMA scores but not with Functional Independence Measure scores.

Kraft, Fitts, and Hammond (1992) tested functional improvement in the upper limb of 22 patients with stroke who received either EMG-initiated electrical stimulation of wrist extensors, low-intensity electrical stimulation of wrist extensors combined with voluntary contractions, proprioceptive neuromuscular facilitation exercises, or no treatment. During the course of treatment, FMA scores of patients receiving proprioceptive neuromuscular facilitation improved 18%, patients receiving low-intensity electrical stimulation of wrist extensors combined with voluntary contractions improved 25%, and patients receiving EMG-initiated electrical stimulation of wrist extensors improved 42%. FMA improvement of the treated groups was significant from pre-treatment to post-treatment, and the improvement was maintained at three-month and nine-month follow-up sessions. In contrast, the control group showed no significant change in FMA scores or grip strength.

Feyes, deWeerdt, Seltz, Steck, Spichinger, and Vereeck (1998) randomized 100 patients with stroke to an experimental group that received sensorimotor stimulation for 6 weeks or to a control group. Patients were evaluated before, during, and after the intervention period and at 6 and 12 months after stroke. Only scores on the FMA showed group differences at follow-up. Scores on the Action Research Arm Test and the Barthel Index did not show group differences.

Duncan et al. (1998) randomized 20 patients with mild or moderate stroke who had completed inpatient rehabilitation to receive a home-based exercise program or usual care. The experimental group demonstrated more improvement in the FMA Motor domain upper and lower extremity subscores than did the usual care group. However, the differences in motor recovery were only significant for the Motor domain lower extremity (Motor domain upper extremity mean change in score = 8.4 versus 2.2; Motor domain lower extremity mean change in score = 4.7 versus -0.9).

Malouin et al. (1994) tested 32 patients with the FMA and the Motor Assessment Scale on two consecutive days and found that the FMA distinguished between patients with minimal recovery better than the Motor Assessment Scale. Adequate to excellent negative correlations between score differences and levels of recovery (upper extremity r = -0.50 and lower extremity r = -0.69) were found, indicating that the largest differences between the two measures were in earlier stages of recovery or among more severely affected individuals.

Platz et al. (2005) tested the construct validity of the FMA upper extremity items (including items from the Motor function, Sensation and passive Joint motion/Joint pain subscores), the Action Research Arm Test, the Box and Block Test and the Motricity Index, using the Spearman correlation coefficient, in patients with upper limb paresis either from stroke (n=37), multiple sclerosis (n=14) or from traumatic brain injury (n=5). Excellent correlations were found between the FMA and the Action Research Arm Test (r = 0.93), the Box and Block Test (r = 0.92), and the Motricity Index (r = 0.86). The FMA was also correlated with more general measures of impairment and activity limitation, including the Ashworth Scale, the Hemispheric Stroke Scale and the modified Barthel Index. An excellent correlation was found between the FMA and the Hemispheric Stroke Scale (r = -0.69), and an adequate correlation was found between the FMA and the Ashworth Scale (r = -0.42). Only a poor correlation was found between the FMA and the modified Barthel Index (r = 0.09).
Note: Correlations were negative because a high score on the FMA indicates normal performance, where as a low score on the Hemispheric Stroke Scale or the Ashworth Scale indicates normal performance.

Fugl-Meyer and Jaasko (1980) compared the FMA Motor domain to performance on activities of daily living scale in 64 patients with stroke more than 6 months after hospital discharge. The activities of daily living scale was comprised of 52 items examining independence in feeding, hygiene, dressing, locomotion, housework, and psychosocial functioning. Excellent correlations were found between the degree of motor impairment as measured by the FMA and the activities of daily living scale (activities of daily living total score r = 0.75; hygiene r = 0.89; locomotion r = 0.76; feeding r = 0.72; and dressing r = 0.76).

Convergent:
Mao et al. (2002) compared the psychometric properties of the Berg Balance Scale, the modified Balance subscore of the FMA, and the Postural Assessment Scale for Stroke Patients, in 123 patients with stroke followed up prospectively 14, 30, 90, and 180 days after stroke onset. Excellent correlations were found using the Spearman correlation coefficient between the scores of the FMA Balance domain and the scores of the Berg Balance Scale at all four time points (ranging from r = 0.86 to r = 0.89), indicating excellent convergent validity.

Rabadi and Rabadi (2006) examined 104 inpatients with acute stroke at a rehabilitation unit. The Action Research Arm Test, the Motor domain upper extremity subscore of the FMA, the NIH Stroke Scale, the Functional Independence Measure total score, and Functional Independence Measure activities of daily living subscore were administered. Using the Spearman rank correlation coefficient, the Action Research Arm Test and the Motor domain upper extremity subscore of the FMA correlated excellently with one another, both at admission (r = 0.77) and at discharge (r = 0.87). The Motor domain upper extremity subscore of the FMA and the Functional Independence Measure activities of daily living subscore at the time of admission were adequately correlated (r = 0.54).

Arsenault et al. (1988) treated 62 patients with hemiplegia with the Bobath approach to treatment for a period of three months. During this time they were evaluated on three occasions. Using Spearman’s Rho, the FMA correlated excellently with the Bobath Assessment of upper extremity pre- (r = 0.73) and post-rehabilitation (r = 0.85), with both measures showing change across time periods.

Dettmann et al. (1987) administered the FMA and the Barthel Index to 15 patients with stroke. Patients were assessed at an average of 2 years post-stroke. Pearson correlations between FMA and Barthel Index scores were excellent for the FMA Total score (r = 0.67), the FMA Motor subscore (r = 0.74), the Motor domain upper extremity subscore (r = 0.75), and the Balance subscore (r = 0.76). In this study, the authors measured walking performance using photographic analysis of gait pattern and velocity, and postural stability while standing on a force platform. The FMA Motor domain lower extremity subscore correlated well with most of the gait measurements, and the FMA Balance subscore correlated well with the stability index. The Sensation subscore did not correlate significantly with any of these measures of gait or upright stability.

Shelton, Volpe, and Reding (2000) compared the FMA to the Functional Independence Measure in 172 inpatients in a stroke rehabilitation hospital within 90 days of stroke. Total FMA scores were highly correlated with total Functional Independence Measure scores (r = 0.63). The correlation between the FMA Motor domain upper extremity subscore and the Functional Independence Measure self-care scores was excellent (r = 0.61), as was the correlation between the Motor domain lower extremity subscore of the FMA and the Functional Independence Measure mobility score (r = 0.74).

Lin et al. (2004) examined the convergent validity of the FMA Sensation subscore using Spearman’s Rho in 176 patients with stroke. The FMA Sensation subscore was poor to adequately correlated with the Barthel Index (ranging from r = 0.38 to r = 0.53). FMA Sensation subscore was also poor to adequately correlated to the FMA Motor domain subscore at different post-stroke stages of recovery, indicating low to moderate convergent validity (ranging from r = 0.31 to r = 0.44).

Known groups:
Poole and Whitney (1988) administered the Motor Assessment Scale and the FMA to 30 patients with hemiplegia. The FMA lower extremity subscore was able to distinguish between patients who needed assistance in walking better than gait speed at speeds less than 0.34 meters/second (r = 0.62).

Bernspang, Asplund, Eriksson, and Fugl-Meyer (1987) administered the FMA to 109 patients within two weeks of having an acute stroke. The FMA was found to distinguish between three levels of self-care ability (dependent, partly dependent, and independent).

Responsiveness

Mao et al. (2002) reported a significant change in the modified FMA Balance subscore between times of assessment (14, 30, 90 and 180 days post-stroke). Effect sizes were large in the interval between 14 and 30 days (ES = 0.82) and weakened the further one moved through time from the stroke event (90-180 days was small, ES = 0.33). The overall effect size (14-180 days) was large (ES = 1.14). When patients were grouped by level of stroke severity, the highest overall FMA Balance subscore effect size among severe stroke patients was an ES = 1.57.

Van der Lee et al. (2001) examined 22 patients with chronic stroke who underwent intensive forced-use treatment to improve upper extremity function. A responsiveness ratio (the ratio of the mean change after the experimental intervention and the standard deviation of the mean change during the baseline period) of 2.03 was reported for the Action Research Arm Test and 0.41 for the FMA Motor score. A higher responsiveness ratio indicates greater responsiveness, suggesting that the Action Research Arm Test is more responsive than the FMA to improvement in upper extremity function in response to a forced-use treatment intervention. This was also reflected by the number of patients who improved more than the upper limit of agreement on the Action Research Arm Test during the intervention period, 12 (54.5%), in comparison to only 2 (9.1%) on the FMA, indicating further that the Action Research Arm Test is more responsive to change than the FMA.

Rabadi and Rabadi (2005) assessed the responsiveness of the Action Research Arm Test and the FMA in evaluating recovery of upper extremity function in 104 inpatients with acute stroke. The mean change in score from admission to discharge was 10 ± 15 for the Action Research Arm Test and 10 ± 13 for the FMA Motor score. The responsiveness to change as measured by the standard response mean (SRM) was moderate for both the Action Research Arm Test (SRM = 0.68) and for the FMA Motor score (SRM = 0.74).

Lin et al. (2004) examined the responsiveness of the FMA Sensation subscore using the standardized response mean (SRM) in 176 patients with stroke. Small (90-180 days SRM = 0.27; 14-30 days SRM = 0.42; 30-90 days SRM = 0.43) to moderate (14-180 days SRM = 0.67) responsiveness was reported for the FMA Sensation subscore.

Wood-Dauphinee et al. (1990) compared the FMA to the Barthel Index in 167 patients with stroke assessed shortly after admission to the hospital and 5 weeks later. Using Pearson correlation coefficients, the correlation between mean change scores for FMA Upper and lower extremity Motor subscores and total Barthel Index scores was adequate (r = 0.57). Small effect sizes were reported for the FMA Motor scale from admission to 5 weeks post-stroke (0.2 for upper extremity, 0.19 for lower extremity, 0.33 for balance ability, and 0.24 for Total score). The results of this study suggest that the FMA Motor scale has small responsiveness.

Duncan, Lai and Keighley (2000) examined the speed and extent of recovery in 459 patients over the first 6 months following stroke. Spontaneous improvement was observed for mild, moderate, and severe strokes as measured by the FMA. Within the first month post-stroke, maximum recovery was achieved, and began to plateau around 6 months post-stroke. The recovery curves of the FMA upper extremity and lower extremity Motor subscales’ corresponded to the recover curves of the Barthel Index and the NIH Stroke Scale. If recovery is defined by achieving an FMA score > 90, then 36.8% of the patients in this study were considered to have recovered.

Shelton et al. (2000) found a moderate correlation between the change in Total FMA score with the change in Total Functional Independence Measure score (r = 0.44). Linear regression analysis demonstrated that with every 24-point increase in Functional Independence Measure score, a 10-point increase in FMA Motor score is observed, indicating that improvement in motor impairment is associated with significant functional recovery. Change in both upper and lower extremity FMA Motor subscores correlated poorly with the change in self-care and mobility Functional Independence Measure subscores (r = 0.23 and r = 0.18, respectively).

Hsueh et al. (2009) examined the responsiveness of the FMA, the Stroke Rehabilitation Assessment of Movement (STREAM) and their shortened versions in 50 clients with chronic stroke. Participants were assessed at two points in time: at admission and at discharge from a rehabilitation program. Both the STREAM and the FMA shortened versions demonstrated a moderate effect size of 0.53 and 0.51, while the STREAM and FMA demonstrated a small effect size of 0.45 and 0.38, respectively.

References
  • Arsenault, A. B., Dutil, E., Lambert, J., Corriveau, H., Guarna, F., Drouin, G. (1988). An evaluation of the hemiplegic subject based on the Bobath approach. Part III. A validation study. Scand J Rehabil Med, 20(1), 13-16.
  • Beckerman, Vogelaar, T. W., Lankhorst, G. J., Verbeek, A. L. (1996). A criterion for stability of the function of the lower extremity in stroke patients using the Fugl-Meyer Assessment Scale. Scand J Rehabil Med, 28, 3-7.
  • Berglund, K., Fugl-Meyer, A.R. (1986). Upper extremity function in hemiplegia: A cross validation study of two assessment methods. Scandinavian Journal of Rehabilitation Medicine, 18, 155-157.
  • Bernspang, B., Asplund, K., Eriksson, S., Fugl-Meyer, A. R. (1987). Motor and perceptual impairments in acute stroke patients: effects on self-care ability. Stroke, 18, 1081-1086.
  • Chae, J., Johnston, M., Kim, H., Zorowitz, R. (1995). Admission motor impairment as a predictor of physical disability after stroke rehabilitation. Am J Phys Med Rehabil, 74(3), 218-223.
  • Chae, J., Bethoux, F., Bohine, T., Dobos, L., Davis, T., Friedl, A. (1998). Neuromuscular stimulation for upper extremity motor and functional recovery in acute hemiplegia. Stroke, 29(5), 975-979.
  • Chae, J., Labatia, I., Yang, G. (2003). Upper limb motor function in hemiparesis: Concurrent validity of the arm motor ability test. Am J Phys Med Rehabil, 82, 1-8.
  • Crow, J.L., Harmeling-van der Wel, B.C. (2008). Hierarchical properties of the motor function sections of the Fugl-Meyer Assessment Scale for people after stroke: a retrospective study. Physical Therapy, 88(12), 1554-1567.
  • Dettmann, M. A., Linder, M. T., Sepic, S. B. (1987). Relationships among walking performance, postural stability, and functional assessments of the hemiplegic patient. Amer J Phys Med, 66, 77-90.
  • De Weerdt, W., Harrison, M. A. (1985). Measuring recovery of arm-hand function in stroke patients: a comparison of the Brunnstrom-Fugl-Meyer test and the Action Research Arm test. Physiother Canada, 37, 65-70.
  • Di Fabio, R. P., Badke, M. B. (1990). Relationship of sensory organization to balance function in patients with hemiplegia. Phys Ther, 70(9), 542-548.
  • Duncan, P. W., Lai, S. M. Keighley, J. (2000). Defining post-stroke recovery: implications for design and interpretation of drug trials. Neuropharmacology, 39(5), 835-41.
  • Duncan, P. W., Propst, M., Nelson, S. G. (1983). Reliability of the Fugl-Meyer assessment of sensorimotor recovery following cerebrovascular accident. Phys Ther, 63(10), 1606-1610.
  • Duncan, P., Richards, L., Wallace, D., Stoker-Yates, J., Pohl, P., and Luchies, C. (1998). A randomized, controlled pilot study of a home-based exercise program for individuals with mild and moderate stroke. Stroke, 29, 2055-2060.
  • Duncan, P. W., Goldstein, L. B., Horner, R. D., Landsman P. B, Samsa, G. P., Matchar, D. B. (1994). Similar motor recovery of upper and lower extremities after stroke. Stroke, 25, 1181-1188.
  • Ferraro, M., Demaio, J. H., Krol, J., Trudell, C., Rannekleiv, K., Edelstein, L., Christos, P., Aisen, M., England, J., Fasoli, S., Krebs, H., Hogan, N., Volpe, B. T. (2002). Assessing the Motor Status Score: A Scale for the Evaluation of Upper Limb Motor Outcomes in Patients after Stroke. Neurorehabilitation and Neural Repair, 16(3), 283-289.
  • Feyes, H. M., deWeerdt, W. J., Seltz, B. E., Steck G. A. C., Spichinger, R., Vereeck, L. E. (1998). Effect of therapeutic intervention for hemiplegic upper limb in acute phase after stroke. Stroke, 29, 785-792.
  • Feys, H., Van Hees, J., Bruyninck, F., Mercelis, R., De Weerdt, W. (2000). Value of somatosensory and motor evoked potentials in predicting arm recovery after a stroke. J Neurol Neurosurg Psychiatry,68, 323-331.
  • Finch, E., Brooks, D., Stratford, P. W., Mayo, N. E. (2002). Physical Rehabilitations Outcome Measures. A Guide to Enhanced Clinical Decision-Making (2nd ed.). Canadian Physiotherapy Association, Toronto.
  • Fugl-Meyer, A. R., Jaasko, L., Leyman, I., Olsson, S., Steglind, S. (1975). The post-stroke hemiplegic patient: I. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine, 975(7), 13-31.
  • Fugl-Meyer, A. R. (1980) Post-stroke hemiplegia assessment of physical properties. Scandinavian Journal of Rehabilitation Medicine, 7, 85-93.
  • Fugl-Meyer, A. R., Jaasko, L. (1980). Post-stroke hemiplegia and ADL-performance. Scand J Rehabil Med Suppl. 7, 140-152.
  • Fulk, G. D., Reynolds, C., Mondal, S., & Deutsch, J. E. (2010). Predicting home and community walking activity in people with stroke. Arch Phys Med Rehabil, 91, 1582-1586.
  • Gladstone, D. J., Danells, C. J., Black, S. E. (2002). The Fugl-Meyer Assessment of Motor Recovery after Stroke: A critical review of its measurement properties. Neurorehabilitation and Neural Repair, 16, 232-240.
  • Gowland, C., Van Hullenaar, S., Torresin, W. (1995). Chedoke-McMaster stroke assessment: development, validation and administration manual. Hamilton (ON), Canada: Chedoke-McMaster Hospitals and McMaster University.
  • Gowland, C., Stratford, P., Ward, M. (1993). Measuring physical impairment and disability with the Chedoke-McMaster Stroke Assessment. Stroke, 24, 58-63.
  • Hsieh, Y.W., Hsueh, I.P., Chou, Y.T., Sheu, C.F., Hseih, C.L. & Kwakkel, G. (2007). Development and validation of a short form of the Fugl-Meyer motor scale in patients with stroke. Stroke, 38, 3052-3054.
  • Hsueh, I. P., Hsu, M. J., Sheu, C. F., Lee, S., Hsieh, C. L., Lin, J. H. (2009). Psychometric comparisons of 2 versions of the Fugl-Meyer Motor Scale and 2 versions of the Stroke Rehabilitation Assessment of Movement. Neurorehabil Neural Repair, 22, 737.
  • Hsueh, I. P., Mao, H. F., Huang, H. L., Hsieh, C. L. (2001). Comparisons of responsiveness and predictive validity of two balance measures in stroke inpatients receiving rehabilitation [in Chinese]. Formos J Med, 5, 261-268.
  • Kraft, G. H., Fitts, S. S., Hammond, M. C. (1992). Techniques to improve function of the arm and hand. Archives of Physical Medicine and Rehabilitation, 73, 220-227.
  • Kusoffsky, A., Wadell, I., Nilsson, B. Y. (1982). The relationship between sensory impairment and motor recovery in patients with hemiplegia. Scand J Rehabil Med, 14(1), 27-32.
  • Lin, J-H., Hsueh, I-P., Sheu, C-F., Hsieh, C-L. (2004). Psychometric properties of the sensory scale of the Fugl-Meyer Assessment in stroke patients. Clinical Rehabilitation, 18, 391-397.
  • Lin, F. M., Sabbahi, M. (1999). Correlation of spasticity with hyperactive stretch reflexes and motor dysfunction in hemiplegia. Arch Phys Med Rehabil, 80(5), 526-530.
  • Malouin, F., Pichard, L., Bonneau, C., Durand, A., Corriveau, D. (1994) Evaluating motor recovery early after stroke: comparison of the Fugl-Meyer Assessment and the Motor Assessment Scale. Arch Phys Med Rehabil, 75(11), 1206-1212.
  • Mao, H. F., Hsueh, I. P., Tang, P. F., Sheu, C. F., Hsieh, C. L. (2002). Analysis and comparison of the psychometric properties of three balance measures for stroke patients. Stroke, 33, 1022-1027.
  • Nadeau, S., Arsenault, A. B., Gravel, D. Bourbonnais, D. (1999). Analysis of the clinical factors determining natural and maximal gait speeds in adults with a stroke. American Journal of Physical Medicine & Rehabilitation, 78(2), 123-130.
  • Nilsson, L., Carlsson, J., Grimby, G., Nordholm, L. (1998). Assessment of walking balance and sensorimotor performance of hemiparetic patients in the acute stages after stroke. Physiother Theory Pract, 14, 149-157.
  • Platz, T., Pinkowski, C., van Wijck, F., Kim, I-H., di Bella, P., Johnson, G. (2005). Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer Test, Action Research Arm Test and Box and Block Test: a multicentre study. Clinical Rehabilitation, 19, 404-411.
  • Poole, J.L. & Whitney, S.L. (1988). Motor Assessment Scale for stroke patients: Concurrent validity and interrater reliability. Archives of Physical Medicine and Rehabilitation, 69, 195-197.
  • Poole, J. L., Whitney, S. L. (2001). Assessments of Motor Function Post Stroke: A Review. Physical & Occupational Therapy in Geriatrics, 19(2), 1-22.
  • Rabadi, M. H., Rabadi, F. M. (2006). Comparison of the Action Research Arm Test and the Fugl-Meyer Assessment as measures of upper-extremity motor weakness after stroke. Arch Phys Med Rehabil, 87, 962-966.
  • Sanford J, Moreland J, Swanson LR, Stratford PW, Gowland C. (1993). Reliability of the Fugl-Meyer assessment for testing motor performance in patients following stroke. Phy Ther, 73, 447-54.
  • Shelton, F. N. A. P., Volpe, B. T., Reding, M. J. (2000). The effect of motor impairment on disability
  • following stroke [abstract]. Stroke, 31(1), 291.
  • Sonde, L., Gip, C., Fernaeus, S.E., Nilsson, C.G., & Viitanen, M. (1998). Stimulation with low frequency (1.7 Hz) transcutaneous electric nerve stimulation (low TENS) increase motor function of the post-stroke paretic arm. Scandinavian Journal of Rehabilitation Medicine, 30, 95-99.
  • Sullivan, K.T., Tilson, J.K., Cen, S.Y., Rose, D.K., Hershberg, J., Correa, A., et al. (2011). Fugl-Meyer Assessment of sensorimotor function after stroke. Standardized training procedure for clinical practice and clinical trials. Stroke, 42, 427-432.
  • Teasell, R., Foley, N. C., & Salter K. (2011). EBRSR: Evidence-Based Review of Stroke Rehabilitation. 13th ed. London (ON): EBRSR.
  • Tyson, S., DeSouza, L. (2002). A systematic review of methods to measure balance and walking post-stroke. Part 1: Ordinal Scales. Physical Therapy Reviews; 7, 173-186.
  • Woodbury, M.L, Velozo, C.A., Richards, L.G., Duncan, P.W., Studenski, S. & Lai, S. (2008). Longitudinal stability of the Fugl-Meyer Assessment of the Upper Extremity. Archives of Physical Medicine and Rehabilitation, 89, 1563-1569.
  • Wood-Dauphinee, S. L., Williams, J. I., Shapiro, S. H. (1990). Examining outcome measures in a clinical study of stroke, Stroke, 21, 731-739.
  • van der Lee, J. H., Beckerman, H., Lankhorst, G. J., Bouter, L. M. (2001). The responsiveness of the Action Research Arm Test and the Fugl-Meyer Assessment Scale in chronic stroke patients. Journal of Rehabilitation Medicine, 33(3), 110-113.