- Email: [email protected]
Dynamic muscle strength training in stroke patients: Effects on knee extension torque, electromyographic activity, and motor function
419
Dynamic Muscle Strength Training in Stroke Patients: Effects on Knee Extension Torque, Electromyographic Activity, and Motor Function Margareta Engardt, PhD, RPT, Evert Knutsson, MD, Phi), Margareta Jonsson, RPT, Mats Sternhag, RPT ABSTRACT. Engardt M, Knutsson E, Jonsson M, Sternhag M. Dynamic muscle strength training in stroke patients: effects on knee extension torque, electromyographic activity, and motor function. Arch Phys Med Rehabil 1995;76:419-25. • The effects of training with isokinetic maximal voluntary knee extensions were studied in stroke patients. Two groups of 10 patients each trained twice a week for 6 weeks. One group trained exclusively eccentric movements and the other exclusively concentric movements. The effects were evaluated from the following tests before and after the training period. The maximal voluntary strength in concentric and eccentric actions of the knee extensor and flexor muscles was recorded together with surface electromyography at constant velocities of 60, 120, and 180 deg. s -~ on three different days. The body weight distribution on the legs while rising and sitting down was measured with two force plates. The self-selected and maximal walking speeds and the swing to stride ratio of the paretic leg were measured. After the training period, the knee extensor strength had increased in eccentric and concentric actions in both groups (p < .05). The eccentric and the concentric strength in the paretic leg relative to that of the nonparetic leg increased in the eccentrically trained group (p < .05) but not in the concentrically trained group. The restraint of the antagonistic muscles in concentric movements increased after concentric (p < .05) but not eccentric training. A nearly symmetrical body weight distribution on the legs in rising from a sitting position was noted after eccentric (p < .05) but not concentric training. Changes in walking variables were not significantly different between the groups. Eccentric knee extensor training was thus found to have some advantages as compared to concentric training in stroke patients. © 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
In the early phase of rehabilitation after stroke, patients learn to use the nonparetic leg to compensate for the weakness of the paretic leg. When muscle strength with time becomes partially restituted, the patient may continue to use the nonparetic leg more than actually needed because the patient has become used to relying on the nonparetic leg with minor use of the paretic leg in the movements of daily life. This mechanism, called learned disuse, often constitutes a problem in the rehabilitation of stroke patients. Because muscle weakness is one of the most prominent consequences of stroke, z3 it was considered important to determine whether training in order to improve leg muscle strength could have an effect in limiting the learned disuse of the affected leg. Strength training has not been widely used in stroke rehabilitation because it has been believed to interfere with coordination and timing in motor control, 4 but this concept has recently been challenged. 5"6 However, it is From the Department of Physical Therapy (Dr. Engardt), Karolinska Institute, the Department of Clinical Neurophysiology (Dr. Knutsson), Karolinska Hospital, and the Unit of Physical Therapy (Ms. Jonsson, Mr. Sternhag), Karolinska Hospital, Stockholm. Supported by research funds from The 1987 Foundation for Stroke Research, Hagersten, the Funds of the Karolinska Institute, and The Board of Physical Therapy Education, Karolinska Institute, Stockholm, Sweden. Submitted for publication March 21, 1994. Accepted in revised form November 28, 1994. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. This paper was presented in part at the International Congress on Stroke Rehabilitation, Berlin, Germany, November 21-24, 1993. Reprint requests to Margareta Engardt, RPT, PhD, Department of Physical Therapy, Karolinska Institute, S-141 57 Huddinge, Sweden. © 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation 0003-9993/95/7605-301853.00/0
not very well-known how an improvement of muscle strength can best be attained in stroke patients because experiences of forceful training, needed to increase muscle strength, are virtually lacking. With active, isokinetic dynamometers, voluntary movements can be controlled at well-defined constant velocities of angular rotation and different modes of muscle action, (ie, concentric and eccentric). They allow training within a large movement range and supply resistance to the patients' maximal voluntary, contractile tension within the whole range of movement. It gives the possibility of training concentfic and eccentric movements of different velocities with optimal loading of the muscles by adjusting the resistance to the actual strength at all phases of maximal voluntary movements of different modes and speeds. In concentric movements, the antagonists are stretched by the movement. This stretch can activate the exaggerated stretch reflexes in spastic muscles. In patients with spastic paresis (eg, spastic hemiparesis after stroke), the antagonist restraint in concentric movements can become abnormally high because the stretch reflex activation adds to the normal antagonist coactivation. 7 In concentric movements at maximal voluntary effort, the antagonist restraint increases with the movement velocity due to the velocity dependence of the exaggerated stretch reflexes.~ This leads to strong antagonist restraint in concentric movements of high velocities. Thus, training of maximal voluntary concentric movements may imply an unfavorable increase of antagonist restraint as suggested after studies of concentric training in patients with spastic paresis. 7 In contrast, there is no antagonist restraint in eccentric movements. Instead, stretch reflexes may act in Arch Phys Med Rehabil Vol 76, May 1995
420
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
synergy with the voluntary commands, s It leads to high tension, which is favorable in strength training. However, it is not clear whether strength training has an effect on motor functions of daily life in stroke patients. Will strength training of the quadriceps muscles performed in eccentric actions have any effect on the concentric action of this muscle, needed in rising to standing from a sitting position? Neither is it known whether training of its concentric action will have any prominent effects on its eccentric action in sitting down. Will strength training of the quadriceps muscles have influence on walking velocity? Therefore, the paretic leg was trained with concentric movements in one group and eccentric movements in another to enable comparison of the effects on voluntary strength and on motor functions of daily life. PATIENTS
AND METHODS
Patients
Twenty adult men and women subjects with hemiparesis secondary to a cerebrovascular accident participated in the study. All patients were ambulatory with (n = 12) and without (n = 8) assistive devices. The patients were allocated to two groups: one group exclusively for eccentric (n = 10) and the other exclusively for concentric (n = 10) training. The allocation was performed according to the patients' age, sex, affected side, mean duration of lesion, and to clinical symptoms (ie, muscle tone and sensibility) to minimize imbalance between the two training groups. Physical performance of the lower extremities was assessed according to Fugl-Meyer and associates. 9'~° A threepoint ordinal scale (0 to 2), for gradings of motor function, balance, light touch and position sense, passive range of motion, and occurrence of joint pain of the lower extremities, was used. Scoring is cumulative from 0 to 100 points. Motor function of sitting to standing and of walking, respec= fively, was tested using the Motor Assessment Scale (MAS), designed by Carr and associates. 11'12Two seven-point ordinal scales (0 to 6), one for each function, were used. Muscle tone of the quadriceps and hamstring muscles was estimated from the resistance to manual passive joint movements using a five-point ordinal rating scale with 0 indicating normal muscle tone and 4 severe hypertone (extremity fixed in flexion or extension), according to Ashworth. 13 Table 1 shows the patients' characteristics and scored assessments, which were not different between the groups before the start of the strength training. Torque
The torque of isokinetic maximal voluntary concentric and eccentric knee extensor and flexor actions of the paretic and the non-paretic leg was measured with a KIN-COM 500H dynamometer, a The patients sat on the dynamometer table with its back support inclined backwards to an angle of 10° from the vertical plane. TM They kept their arms folded in front of the chest during the tests. For fixation of the patients' thigh of the leg to be examined, one strap was positioned just inferior to the anterior superior iliac spines and another over the thigh 20cm proximal to the superior border of the patella. The axis of Arch Phys Med Rehabil Vol 76, May 1995
Table 1: Characteristics and Assessments of the Stroke Patients: Mean ± SD, Median, Ranges, and Numbers
Age (year) Weight (kg) Height (cm) Months since onset (n) Physical performance (F-M)* Sit-to-stand (MAS) t Walking (MAS) t Muscle tone (Ashworth)* Sensibility~ Affected side (left/right) Female/male Walking aids (with/without)
Eccentrically Trained (n = 10)
Concentrically Trained (n = 10)
62.2 ± 7.6 72.9 ± 14.3 173 ± 9.9 26.5 ± 10.3 81 (68-97) 2 (2-6) 5 (4-6) ---4 (n = 6) >4(n=4) ---8 (n = 2) >8(n=8) 7/3 3/7 7/3
64.6 ± 6.2 81.6 ± 10.9 174 ± 9.9 27.8 ± 12.0 78 (70-86) 2 (2-2) 4 (3-5) --_4 (n = 8) >4(n =2) ---8 (n = 1) >8(n=9) 5/5 2/8 5/5
* F-M, Fugl-Meyer score (maximum, 100 points). * MAS, motor assessment scale (maximum, 6 points). * Ashworth scale (maximum spasticity, lower limb = 12 points). F-M, Fugl-Meyer score (maximum, 12 points).
rotation of the dynamometer was aligned as well as possible to the slightly changing rotation axis of the knee joint. The pad of the lever arm of the dynamometer was placed with its lower end 2cm above the patients' medial malleolus. In the test, the torque caused by the weight of the leg and foot was first measured for gravitational corrections. After a few pretest trials to get accustomed to the dynamometer and after movements at submaximal efforts to warm up, series of repeated measurements of torque and electromyographic (EMG) activity were made during isokinetic maximal voluntary movements at three velocities, 60, 120, and 180 deg-s -1. At each velocity, data from three repetitions of a concentric movement followed by an eccentric movement were averaged and stored. Measurements were accepted only when three repetitions gave similar torque curves with torque variations less than 15%. The range of movement was 20 to 90 ° except in one patient in whom the range was 30 to 90 ° due to a smaller range of his voluntary knee movements. Between the tests at different velocities, periods of rest for 3 minutes were allowed, and between the individual repetitions there were 20-second pauses for rest. The voluntary strength was tested on the nonparetic side on one occasion and on the paretic side on three different days before and after the training period. The average value of the torque (Nm) obtained during movement through a selected range, 31 to 80 ° of knee flexion, was used to compare the maximal muscle strength at different movement velocities and different modes of muscle action. The movement range selected was the largest with constant velocity in all movements used in the tests. It was obtained by excluding the phases of acceleration and deceleration in isokinetic knee extensions and flexions at the highest velocity (180 deg-s -1) in which these phases are largest. In one patient, a smaller range, 45 to 80 °, had to be used owing to reduced range of voluntary movements. (EMG) The EMG signals were recorded with surface electrodes (Medicotest Type A-10-VS) b over the quadriceps and the
Eleetromyography
STRENGTH TRAINING IN STROKE PATIENTS, Engardt hamstring muscle groups. Skin preparation was performed by 95% ethanol washing and rubbing of the skin over the areas of the electrode placements. Two pairs of electrodes were placed as follows: quadriceps muscles: across the thigh, one third of the thigh length from the upper border of the patella and one fifth of the thigh circumference apart, and centered to a line between patella and the midpoint of ligamentum inguinale; hamstring muscles: along lines from tuber ossis ischii to the distal tendons of biceps femoris and semitendinosus at a distance from fossa poplitea 40% of the thigh length.~5 With these electrode positions, the total action of the knee extensor and flexor muscles is usually mirrored quite well. The EMG signals were amplified (1,000×) in miniature preamplifiers (Oxford HDX-82) a placed closely to the surface electrodes and fed to the standard EMG processing unit of the KIN-COM dynamometer system (KC/EMG)." In this, the EMG signals are integrated by rectification and bandpass filtering (20 to 1,000Hz) in a Paynter filter, followed by low-pass filtering (100Hz). 1637 The voltage signal thus obtained indicates the level of EMG activity. It was sampled at 100Hz using a 12 bit A/D converter. The integrated EMG was averaged for the same movement range as the one used to determine average torque.
10_ 98_ 7_
z5_
o4_
..J
i'
2 1 0
Body Weight Distribution on the Legs in Rising and in Sitting D o w n The vertical floor reaction forces under each foot during rising and sitting down were recorded with two force platforms. ~8 The patients sat in a standardized position on an adjustable chair with a back support. The seat height was set to the patient's knee height, determined as the distance from the lateral knee joint line to the floor, with the tibia perpendicular to the floor. The trunk was upright and the thighs supported for 3 of the femoral length. The patient placed one foot on each platform. The patients were instructed to stand up, to stand for 1 minute, and to sit down again. The instructions were "Please stand up, as you usually d o " and after 1 minute of standing, "Please sit down, as you usually do". A few pretest trials were allowed for the patients to become accustomed to the testing procedure. Strain gauge transducers fixed to each one of the platforms were used to record the vertical force under each foot during the test movement. The body weight distribution was estimated from the force load on the platforms during the test movements (ie, the time integral of the vertical forces under each foot). It was calculated from the area under the force curve from start to end of rising and sitting down, respectively (fig 1). The body weight distribution was defined as the ratio between the time integral of the vertical forces of the paretic and of the nonparetic leg. The means from three repeated tests of rising and of sitting down were calculated. The ratio of the impulses was expressed in percentage of body weight (%BW) supported by each leg. TM Gait The patients walked 30m under the following condition: (1) at their self-selected walking speed and (2) as fast as
421
0
I' 1
2
I 3
4
I 5
TIME (s) Fig 1--Areas under force-time curves (Ns) indicating the vertical forces of the nonparetic and the paretic leg, respectively, of a stroke patient when rising and when sitting down. Nonparetic leg, $~; paretic leg, i possible on an ordinary level floor outside the laboratory area. The gait was evaluated by measuring the mean walking speed and mean duration of the swing phase in percentage of the stride. One pair of pressure sensitive foot switches (Cefar) c was used to indicate the gait cycles of the paretic leg. One foot switch was taped to the heel of the shoe and the other to the sole of the shoe under the metatarsal head of the big toe of the paretic leg. The signals were recorded telemetrically (Medenik Biotelemetry System IC-600), d transmitted and fed to receivers, connected to ACamplifiers (Grass 7 P1F), e and printed on a polygraphic recorder. The collected data were processed in a computer-controlled digitizer system. For each gait cycle, the foot floor contact and the toe off of the paretic leg were marked on the stridecurve with a digitiser stylus for calculating the swing phase duration of the stride. 19
Training Device and Training Procedure Twice a week for 6 weeks, 10 patients trained the paretic leg exclusively with isokinetic maximal voluntary eccentric knee extensor actions and 10 trained exclusively with isokinetic maximal voluntary concentric knee extensor actions. A dynamic dynamometer (KIN-COM 125H) a controlled the movements. The patients sat as described under measurements of the torque (see previous section). After each active Arch Phys Med Rehabil Vol 76, May 1995
422
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
Table 2: Mean Torque (Nm) of Maximal Voluntary Eccentric and Concentric Knee Extensor Actions of the Paretic Leg at Three Different Constant Velocities Before and After the Training Period Eccentrically Trained (n = 10) Velocity Eccentric actions
Concentric actions
60 120 180 60 120 180
d e g ' s -j deg-s ~ d e g . s -~ d e g . s -~ deg-s J d e g . s -j
Before 106.3 113.8 114.4 62.4 47.2 35.3
_+ 12.9 ,+ 15.1 _+ 15.4 .+ 7.9 .+ 6.6 .+ 5.3
Concentrically Trained (n = 10)
After 136.6 142.5 142.3 79.1 59.1 44.2
p
+_ 15.8 .+ 17.8 ,+ 19.0 ,+ 9.5 ,+ 7.5 -4- 6.6
Before
<.01 <.01 <.01 <.05 <.05 <.05
109.4 112.8 112.4 61.8 44.1 29.8
After
,+ 11.0 _ 10.5 _+ 10.3 .+ 7.6 .+ 5.8 .+ 4.2
124.2 130.4 130.1 73.9 57.0 43.2
,+ ,+ .+ .+ .+ .+
p
11.2 11.0 10.3 8.0 6.1 4.7
<.01 <.05 <.05 <.05 <.01 <.01
Mean _+ SEM are given, p was calculated using Wilcoxon's signed rank test.
movement, the dynamometer moved the leg passively back to start position for the next active movement. The sessions started with 3 sets of 10 submaximal repetitions at 60deg. s -1, in the same mode as allotted for training, as a warm-up session. The range of motion of the knee joint was 10 to 100 ° except in a few patients, where the range was reduced and set to the patients' maximal capacity. Thereafter, the patients were asked to make a maximal contraction at 60deg. s -1, and a target curve was created. The training session consisted of repeated sets of 10 repetitions at angular velocities of 60, 120, 180, 120, 60, 120, 180deg. s -1, or more, with a 60-second rest interval after each selected velocity. To check that maximal effort was used the torque curves were superimposed on the first target curve on the monitor. The number of sets of repetitions increased similarly during the training period up to a maximum of 15 sets per training session. Each training session was followed by two 20-second maximal stretchings of the knee extensor muscles given by the physical therapist responsible for the training. Data Analyses
Differences within and between groups were tested for significance by the Wilcoxon's signed rank test and the Mann-Whitney U test. Nominal data were treated by the Chi-square exact test. 2° The level of significance chosen was 5%. RESULTS There were significant changes in mean maximum strength within both of the groups after training, as shown in table 2. In the eccentrically trained group, the mean changes were within the range of 30% to 25% for eccentric and concentric actions. In the concentrically trained group, the mean changes of torque were low (17% to 13%) in eccentric actions and high in concentric actions (25% to 57%). No statistically different changes between the two groups were found. Figure 2 shows the mean ___SEM values of the strength in the paretic leg relative to that of the nonparetic leg before and after the training period. Significant changes were found in the relative strength in the eccentrically trained group in both eccentric and concentric actions. The relative strength increased from 70%, 71% and 72% at 60, 120 and 180deg. s -1, respectively, before the training period to 89%, 88% and 86% after the training period in eccentric actions and from 62%, 54% and 50% to 76%, 62% and 60%, respectively, in concentric actions in the patients of the eccentrically Arch Phys Med Rehabil Vol 76, May 1995
trained group. Corresponding figures for the patients in the concentrically trained group were 78%, 76% and 78% before the training period and 85%, 84% and 86% after the training period in eccentric actions and 71%, 58% and 55% before the training period and 68%, 62% and 58% after the training period in concentric actions. The changes in the concentric group were not significant. Figure 3 shows the mean changes of the averaged agonist EMG activity (%) during voluntary knee extensor actions after training. In the eccentrically trained group, the mean changes were within the range of 33% to 24% for eccentric and concentric actions (p < .05). In the concentrically trained group, the mean changes were within the range of 28% to 23% in eccentric actions and 36% to 37% in concentric actions (p < .05). No statistically different changes between the two groups were found. Figure 4 shows the mean changes of the averaged antagonist EMG activity (%) during voluntary knee extensor actions after training. In eccentric actions, there was no change or a decrease of antagonist EMG activity in both the eccentrically and the concentrically trained group. However, in concentric actions, the antagonist EMG activity increased with increasing velocities in the concentrically trained group, whereas there was no change or a decrease in the eccentrically trained group. The mean changes of the antagonist EMG activity were significantly different in the concentrically trained group at the knee angular velocity of 120 and 180deg • s- 1compared with those of the eccentrically trained group. A o~
Eccentrically Trained ]
j
j
j
.
~
B
j
80
80
60
60
~o 40
40
~
20
o
.
20 0
-180
-120
-60
Eccentric
Concentrically Trained
60
120 Concentric
Velocity (deg/s)
180
0
.Before [ ] After
-le0
-120
-60
Eccentric
60
120 180 Concentric
Velocity (deg/s)
Fig 2 - - T o r q u e of maximal voluntary eccentric and concentric knee extensor actions in the paretic leg relative to that of the nonparetic leg at three different constant angular velocities in (A) the eccentrically (n = 10) and (B) the concentrically trained group (n = 10) before and after 6 weeks of isokinetic training. Values are mean +__SEM. P was calculated using the Wilcoxon's signed rank test. *p < .05, **p < .01.
STRENGTH TRAINING IN STROKE PATIENTS, Engardt Voluntary knee extensor action
423
Rising
Sitting down
A
o~
60
---
_>, ._> ~
ECC. tr.
m
60
40
I$
ro
50
-:-- ~
40
:~ -~
30
.-~ ~ c-,
20
o~,^.^
W to
~
20
o or--
t-0
0
I
m
-180-120-60
|
60
Eccentric
120
180
Concentric
Velocity (deg/s) Fig 3 - - C h a n g e s of agonist EMG activity during maximal voluntary eccentric and concentric knee extensor actions at three different constant angular velocities in the eccentrically (n = 10) and the concentrically (n = 10) trained group after 6 weeks of isokinetic training. Values are mean _+ SEM. P was calculated using the Mann-Whitney U test. The changes in the two groups were not significantly different.
Figure 5 shows the body weight distributed on the paretic leg (%BW) in rising and in sitting down before and after the training period. A significant improvement towards symmetrical body weight distribution on the legs in rising was noted only within the eccentrically (p < .01) trained group. Voluntary knee extensor action
A
v
t
g" ._>
so
w
0
40
IJ.I
"F
O c~
[]
ECC. tr.
•
CON. tr.
7
20
e-
~
0
o
" t-O
10
nq o
r"
B
-20
|
-180 -120
|
-60
|
60
120
180
Eccentric Concentric Velocity (deg/s) Fig 4 - - C h a n g e s of antagonist EMG activity during maximal voluntary eccentric and concentric knee extensor actions at three different constant angular velocities in the eccentrically (n = 10) and the concentrically (n = 10) trained group after 6 weeks of isokinetic training. Values are mean _+ SEM. P was calculated using the Mann-Whitney U test. The changes in the two groups were not significantly different except in concentric actions at 120 and 180deg. s -1. *p < .05.
0 Ecc.tr.
Con.tr.
Ecc.tr.
Con.tr.
Fig 5 - - B W distribution on the paretic leg in percent of the total body weight in rising and sitting down in the eccentrically (n = 10) and the concentrically trained group (n = 10) before and after 6 weeks of training. Values are mean _+ SD. P was calculated using the Mann-Whitney U test. *p < .05.
The change in the eccentrically trained group differed significantly (p < .05) from that of the concentrically trained group. No improvements of weight distribution on the legs were observed in sitting down in any of the two training groups. Significant changes were noted in walking speed and in duration of swing phase within the concentrically trained group (p < .05). The patients walked faster but loaded their paretic leg for a shorter period of the walking cycle after training. However, the changes in the two groups were not significantly different, as shown in table 3. DISCUSSION The present study shows that eccentric as well as concentric training gave a considerable increase of knee extensor strength after six weeks training. The increase in strength was related to enhanced activation of the agonists as judged from the mean changes of the averaged EMG activity. The changes in the averaged EMG agonist activity were proportional to the changes of torque in eccentric as well as in concentric actions in both groups. This indicates that the strength increase was caused by neural factors. The intensive loading through the isokinetic training might activate previously dormant fibers. 2~ The increase in strength production might be caused by an increased ability to raise motoneuron activation during voluntary effort. 22 This is in accordance with results after training in healthy subjectsY 27 As shown in the present study, there were significant improvements in the relative strength of the paretic leg in eccentric as well as concentric actions after eccentric training but not after concentric training. In an earlier study of patients with spastic paresis, it was shown that there may be considerable restraint to maximal voluntary concentric movements because the spastic reflexes in the antagonists may be facilitated by strong voluntary effort. ~ The repeated forceful voluntary movements of concentric training will Arch Phys Med Rehabil Vol 76, May 1995
424
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
Table 3: Walking Speed (m/s) and Swing Phase Duration (% Gait Cycle) of the Paretic Leg at Self-Selected (A) and Fast Speed (B) in the Eccentrically and the Concentrically Trained Group Before and After Training Eccentrically Trained (n = 10) Before Walking A (m/s) Walking B (m/s) Swing A (% stride) Swing B (% stride)
.81 1.0 43.9 45.1
± .18 ± .25 _+ 6.4 ± 5.9
After .84 1.0 43.9 46.6
_+ .24 _+ .26 + 4.6 ± 4.4
Concentrically Trained (n = 10) p ns ns ns ns
Before .65 .81 42.1 45.1
+_ .2 +_ .29 ± 5.6 ± 6.1
After .73 .91 45.5 47.1
± .25 ± .30 _+ 4.3 ± 6.2
p <.05 <.05 <.05 ns
Note: Changes between the two groups were not statistically different. Mean + SD are given, p was calculated using Wilcoxon's signed rank test and the Mann-Whitney U test.
imply repeated relatively strong reflex activations of the antagonists that may explain the increased antagonist EMG activity after concentric training. An increase in agonist activation caused by training will thus in part be out balanced by increased antagonist reflex activation, which also appears to be an effect of concentric training. In contrast, the antagonists are not activated by stretch reflexes in eccentric training. If stretch reflexes are activated, they are activated in the agonist that is lengthening. Thus, the stretch reflexes will act in synergy with the voluntary effort. These differences may be the reason why eccentric training had a better effect on the voluntary strength. The difference between the effects of eccentric and concentric training appears to be of importance for motor functions of daily life. Thus, the training had different effects on the capacity to load weight symmetrically on the legs during rising up from a sitting posture. After the eccentric training but not after the concentric training, there were significant improvements towards symmetrical body weight distribution in this motor task. It may be a consequence of the larger increase of the strength of the paretic legs relative to the nonparetic after training. It appears likely that the lesser the difference is between the strength of the two lower extremities, the lesser the risk should be of a continued learned disuse of the paretic leg. The action of the quadriceps muscles in rising from a sitting position is a concentric one. The fact that this function was improved after eccentric training implies that the eccentric training had effects not only on the eccentric function trained but also gave a strength increase that could be utilized in concentric actions. No improvements in BW distribution in sitting down were noted after training. However, the distribution of body weight in the patients of the present study was almost equivalent to that of old healthy subjects presented in an earlier study.~8 Thus, an already close to symmetrical relearned motor program of sitting down from standing was not influenced by the improvement of the knee extension torque on average 32 months after stroke. Despite significant improvements in mean knee extension torque, there were no significant changes in the gait parameters except for a significant improvement in walking speed and a prolonged swing phase duration in the concentrically trained group. Mean gait velocity in old healthy men is 1.18m/s (age 67 to 73 years) and 1.45m/s (age 60 to 65 years). 28 As the patients in the eccentrically trained group walked on average with 0.81 at self-selected and 1.0m/s at fastest speeds possible already before training, there might Arch Phys Med Rehabil Vol 76, May 1995
not have been much room or motivation for an increase of the gait velocity. In the concentrically trained group, where the gait velocity was low, the training, however, had an effect. The results of the present study give support for the concept that eccentric training may be more suitable for stroke patients than concentric training. Eccentric training to improve motor functions after stroke was suggested already by Brunnstrtm. 29 However, manual resistance cannot give the same specificity in training. With the isokinetic dynamometer used, movements can be controlled at well-defined movement velocities and maximal contractile tension within a large movement range. Thus, there is a pronounced difference between dynamic training when controlled manually and controlled by a strong active dynamometer. The latter has sufficient capacity to overpower or resist maximal voluntary actions. This implies that eccentric and concentric actions can both be trained at the maximal levels of contractile tension. References 1. Knutsson E, M~rtensson A. Dynamic motor capacity in spastic paresis and its relation to prime mover dysfunction, spastic reflexes and antagonist co-activation. Scand J Rehab Med 1980; 12:93-106. 2. Freund HJ. The pathophysiology of central paresis. In: Struppler A, Weindl A, editors. Electromyography and evoked potentials. Berlin: Springer-Verlag, 1985; 19-20. 3. Langton-Hewer R. Rehabilitation after stroke. Quart J Med 279:65974, 1990. 4. Bobath B. Adult hemiplegia: evaluation and treatment. 2nd ed. London: William Heinemann Medical Books Ltd, 1978:58-63. 5. Bohannon RW. Muscle strength in patients with brain lesions: measurements and implication. In: Harms-Ringdahl K, editor. Muscle strength. New York: Churchill Livingstone, 1993; 187-225. 6. Baker P, Rawicki B. The effect of movement, gait and resistance to stretch of spastic muscle following an isokinetic strengthening program. Proceedings of the Xlth International Congress of World Confederation for Physical Therapy and Chartered Society of Physiotherapy. London: 1991:1009-1111. 7. Knutsson E, M~rtensson A, Gransberg L. The effects of concentric and eccentric training in spastic paresis. Scand J Rehab Med 1992;24(27 Suppl):31-2. 8. Knutsson E. Concentric and eccentric muscle work in spastic paresis. Scand J Rehab Med 1992;24(27 Suppl):16-7. 9. Fugl-Meyer AR, J~iask6 L, Leyman I, Olsson S, Steglind S. The poststroke hemiplegic patient. I. a method for evaluation of physical performance. Scand J Rehab Med 1975;7:13-31. 10. Duncan PW, Propst M, Nelson SG. Reliability of the Fugl-Meyer assessment of sensorimotor recovery following cerebrovascular accident. Phys Ther 1983;63:1606-10. 11. Can" JH, Shepherd RB, Nordholm L, Lynne, D. Investigation of a new motor assessment scale for stroke patients. Phys Ther 1985; 65:175-80. 12. Poole JL, Whitney SL. Motor assessment scale for stroke patients:
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
13. 14. 15. 16. 17.
18. 19.
20.
21. 22.
concurrent validity and interrater reliability. Arch Phys Med Rehabil 1988; 69:195 -7. Ashworth B. Preliminary trial of carisoprodol in multiple sclerosis. Practitioner 1964; 192:540-2. Engardt M, Knutsson E. Dynamic thigh muscle strength after auditory feedback training of body weight distribution in stroke patients. Physiother Theory Practise 1994; 10:103-12. Knutsson E, Richards C. Different types of disturbed motor control in gait of hemiparetic patients. Brain 1979; 102:405-30. Gottlieb GL, Agarwal GC. Filtering of electromyographic signals. Amer J Phys Med 1970;49:142. Gransberg L, Knutsson E, Litton JE. A computer programmed system for the analysis of active and passive isokinetic movements. In: IEEE frontiers of engineering in health care. Proceedings, Institute of Electrical and Electronics Engineers/Medicine and Biology Society, Washington, DC. 1980;292-295. Engardt M, Olsson E. Body-weight bearing while rising and sitting down in patients with stroke. Scand J Rehab Med 1992;24:67-74. Myrenberg M, Lundeberg T, Od6en I. Effect of traction treatment on EMG activity during walking in erecti spinae in patients with low back pain. The Pain Clinic 1990;3:67-73. Mehta CR, Patel NR, Tsiatis AA. Exact significance testing to establish treatment equivalence for ordered categorical data. Biometrics 1984;40:819-25. Perrine JJ. Isokinetic exercise and the mechanical potentials of muscle. JOHPER 1968;5:40-5. Sale DG, MacDougall JD, Upton ARM, McComas AJ. Effect of
23. 24.
25. 26. 27. 28. 29.
425
strength training upon motoneuron excitabilituy in man. Med Sci Sports Exerc 1983; 15:57-62. Barnes W. The relationship of motor-unit activation to isokinetic muscular contraction at different contractile velocities. Phys Ther 1980;60:1152-8. Milner-Brown HS, Stein RB, Lee RG. Synchronization of human motor units: possible roles of exercise and supraspinal reflexes. Electroencephalography Clin Neurophysiol 1975;38:245-54. Rutherford OM, Jones DA. The role of learning and coordination in strength training. Eur J Appl Physiol 1986;55:100-5. HSkkinen K, Komi PV. Electromyographic changes during strength training and detraining. Med Sci Sports Exerc 1983; 15:455-60. Moritani T, De Vries HA. Reexamination of the relationship between the surface integrated electromyogram (IEMG) and force of isometric contraction. Am J Phys Med 1977;57:263-77. Murray MP, Kory R, Clarkson BH. Walking patterns in healthy old men. J Gerontol 1969;24:169-178. Brunnstr~m S. Movement therapy in hemiplegia. New York: Harper & Row, 1970:69.
Suppliers a. Chattecx Corp,, 101 Memorial Drive, PO Box 42887, Chattanooga, TN 37405. b. Medicotest Type A-10-VS. c. CEFAR medical products, S-222 41 Lund, Sweden. d. Medenik AB, S-740 60 t3rbyhus, Sweden. e. Grass Instruments Corporation, Quincy, MA. (in Sweden: Somedic AB, tJstermalmsgatan 27, S-123 42 Farsta, Sweden).
Arch Phys Med Rehabil Vol 76, May 1995
Dynamic Muscle Strength Training in Stroke Patients: Effects on Knee Extension Torque, Electromyographic Activity, and Motor Function Margareta Engardt, PhD, RPT, Evert Knutsson, MD, Phi), Margareta Jonsson, RPT, Mats Sternhag, RPT ABSTRACT. Engardt M, Knutsson E, Jonsson M, Sternhag M. Dynamic muscle strength training in stroke patients: effects on knee extension torque, electromyographic activity, and motor function. Arch Phys Med Rehabil 1995;76:419-25. • The effects of training with isokinetic maximal voluntary knee extensions were studied in stroke patients. Two groups of 10 patients each trained twice a week for 6 weeks. One group trained exclusively eccentric movements and the other exclusively concentric movements. The effects were evaluated from the following tests before and after the training period. The maximal voluntary strength in concentric and eccentric actions of the knee extensor and flexor muscles was recorded together with surface electromyography at constant velocities of 60, 120, and 180 deg. s -~ on three different days. The body weight distribution on the legs while rising and sitting down was measured with two force plates. The self-selected and maximal walking speeds and the swing to stride ratio of the paretic leg were measured. After the training period, the knee extensor strength had increased in eccentric and concentric actions in both groups (p < .05). The eccentric and the concentric strength in the paretic leg relative to that of the nonparetic leg increased in the eccentrically trained group (p < .05) but not in the concentrically trained group. The restraint of the antagonistic muscles in concentric movements increased after concentric (p < .05) but not eccentric training. A nearly symmetrical body weight distribution on the legs in rising from a sitting position was noted after eccentric (p < .05) but not concentric training. Changes in walking variables were not significantly different between the groups. Eccentric knee extensor training was thus found to have some advantages as compared to concentric training in stroke patients. © 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
In the early phase of rehabilitation after stroke, patients learn to use the nonparetic leg to compensate for the weakness of the paretic leg. When muscle strength with time becomes partially restituted, the patient may continue to use the nonparetic leg more than actually needed because the patient has become used to relying on the nonparetic leg with minor use of the paretic leg in the movements of daily life. This mechanism, called learned disuse, often constitutes a problem in the rehabilitation of stroke patients. Because muscle weakness is one of the most prominent consequences of stroke, z3 it was considered important to determine whether training in order to improve leg muscle strength could have an effect in limiting the learned disuse of the affected leg. Strength training has not been widely used in stroke rehabilitation because it has been believed to interfere with coordination and timing in motor control, 4 but this concept has recently been challenged. 5"6 However, it is From the Department of Physical Therapy (Dr. Engardt), Karolinska Institute, the Department of Clinical Neurophysiology (Dr. Knutsson), Karolinska Hospital, and the Unit of Physical Therapy (Ms. Jonsson, Mr. Sternhag), Karolinska Hospital, Stockholm. Supported by research funds from The 1987 Foundation for Stroke Research, Hagersten, the Funds of the Karolinska Institute, and The Board of Physical Therapy Education, Karolinska Institute, Stockholm, Sweden. Submitted for publication March 21, 1994. Accepted in revised form November 28, 1994. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. This paper was presented in part at the International Congress on Stroke Rehabilitation, Berlin, Germany, November 21-24, 1993. Reprint requests to Margareta Engardt, RPT, PhD, Department of Physical Therapy, Karolinska Institute, S-141 57 Huddinge, Sweden. © 1995 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation 0003-9993/95/7605-301853.00/0
not very well-known how an improvement of muscle strength can best be attained in stroke patients because experiences of forceful training, needed to increase muscle strength, are virtually lacking. With active, isokinetic dynamometers, voluntary movements can be controlled at well-defined constant velocities of angular rotation and different modes of muscle action, (ie, concentric and eccentric). They allow training within a large movement range and supply resistance to the patients' maximal voluntary, contractile tension within the whole range of movement. It gives the possibility of training concentfic and eccentric movements of different velocities with optimal loading of the muscles by adjusting the resistance to the actual strength at all phases of maximal voluntary movements of different modes and speeds. In concentric movements, the antagonists are stretched by the movement. This stretch can activate the exaggerated stretch reflexes in spastic muscles. In patients with spastic paresis (eg, spastic hemiparesis after stroke), the antagonist restraint in concentric movements can become abnormally high because the stretch reflex activation adds to the normal antagonist coactivation. 7 In concentric movements at maximal voluntary effort, the antagonist restraint increases with the movement velocity due to the velocity dependence of the exaggerated stretch reflexes.~ This leads to strong antagonist restraint in concentric movements of high velocities. Thus, training of maximal voluntary concentric movements may imply an unfavorable increase of antagonist restraint as suggested after studies of concentric training in patients with spastic paresis. 7 In contrast, there is no antagonist restraint in eccentric movements. Instead, stretch reflexes may act in Arch Phys Med Rehabil Vol 76, May 1995
420
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
synergy with the voluntary commands, s It leads to high tension, which is favorable in strength training. However, it is not clear whether strength training has an effect on motor functions of daily life in stroke patients. Will strength training of the quadriceps muscles performed in eccentric actions have any effect on the concentric action of this muscle, needed in rising to standing from a sitting position? Neither is it known whether training of its concentric action will have any prominent effects on its eccentric action in sitting down. Will strength training of the quadriceps muscles have influence on walking velocity? Therefore, the paretic leg was trained with concentric movements in one group and eccentric movements in another to enable comparison of the effects on voluntary strength and on motor functions of daily life. PATIENTS
AND METHODS
Patients
Twenty adult men and women subjects with hemiparesis secondary to a cerebrovascular accident participated in the study. All patients were ambulatory with (n = 12) and without (n = 8) assistive devices. The patients were allocated to two groups: one group exclusively for eccentric (n = 10) and the other exclusively for concentric (n = 10) training. The allocation was performed according to the patients' age, sex, affected side, mean duration of lesion, and to clinical symptoms (ie, muscle tone and sensibility) to minimize imbalance between the two training groups. Physical performance of the lower extremities was assessed according to Fugl-Meyer and associates. 9'~° A threepoint ordinal scale (0 to 2), for gradings of motor function, balance, light touch and position sense, passive range of motion, and occurrence of joint pain of the lower extremities, was used. Scoring is cumulative from 0 to 100 points. Motor function of sitting to standing and of walking, respec= fively, was tested using the Motor Assessment Scale (MAS), designed by Carr and associates. 11'12Two seven-point ordinal scales (0 to 6), one for each function, were used. Muscle tone of the quadriceps and hamstring muscles was estimated from the resistance to manual passive joint movements using a five-point ordinal rating scale with 0 indicating normal muscle tone and 4 severe hypertone (extremity fixed in flexion or extension), according to Ashworth. 13 Table 1 shows the patients' characteristics and scored assessments, which were not different between the groups before the start of the strength training. Torque
The torque of isokinetic maximal voluntary concentric and eccentric knee extensor and flexor actions of the paretic and the non-paretic leg was measured with a KIN-COM 500H dynamometer, a The patients sat on the dynamometer table with its back support inclined backwards to an angle of 10° from the vertical plane. TM They kept their arms folded in front of the chest during the tests. For fixation of the patients' thigh of the leg to be examined, one strap was positioned just inferior to the anterior superior iliac spines and another over the thigh 20cm proximal to the superior border of the patella. The axis of Arch Phys Med Rehabil Vol 76, May 1995
Table 1: Characteristics and Assessments of the Stroke Patients: Mean ± SD, Median, Ranges, and Numbers
Age (year) Weight (kg) Height (cm) Months since onset (n) Physical performance (F-M)* Sit-to-stand (MAS) t Walking (MAS) t Muscle tone (Ashworth)* Sensibility~ Affected side (left/right) Female/male Walking aids (with/without)
Eccentrically Trained (n = 10)
Concentrically Trained (n = 10)
62.2 ± 7.6 72.9 ± 14.3 173 ± 9.9 26.5 ± 10.3 81 (68-97) 2 (2-6) 5 (4-6) ---4 (n = 6) >4(n=4) ---8 (n = 2) >8(n=8) 7/3 3/7 7/3
64.6 ± 6.2 81.6 ± 10.9 174 ± 9.9 27.8 ± 12.0 78 (70-86) 2 (2-2) 4 (3-5) --_4 (n = 8) >4(n =2) ---8 (n = 1) >8(n=9) 5/5 2/8 5/5
* F-M, Fugl-Meyer score (maximum, 100 points). * MAS, motor assessment scale (maximum, 6 points). * Ashworth scale (maximum spasticity, lower limb = 12 points). F-M, Fugl-Meyer score (maximum, 12 points).
rotation of the dynamometer was aligned as well as possible to the slightly changing rotation axis of the knee joint. The pad of the lever arm of the dynamometer was placed with its lower end 2cm above the patients' medial malleolus. In the test, the torque caused by the weight of the leg and foot was first measured for gravitational corrections. After a few pretest trials to get accustomed to the dynamometer and after movements at submaximal efforts to warm up, series of repeated measurements of torque and electromyographic (EMG) activity were made during isokinetic maximal voluntary movements at three velocities, 60, 120, and 180 deg-s -1. At each velocity, data from three repetitions of a concentric movement followed by an eccentric movement were averaged and stored. Measurements were accepted only when three repetitions gave similar torque curves with torque variations less than 15%. The range of movement was 20 to 90 ° except in one patient in whom the range was 30 to 90 ° due to a smaller range of his voluntary knee movements. Between the tests at different velocities, periods of rest for 3 minutes were allowed, and between the individual repetitions there were 20-second pauses for rest. The voluntary strength was tested on the nonparetic side on one occasion and on the paretic side on three different days before and after the training period. The average value of the torque (Nm) obtained during movement through a selected range, 31 to 80 ° of knee flexion, was used to compare the maximal muscle strength at different movement velocities and different modes of muscle action. The movement range selected was the largest with constant velocity in all movements used in the tests. It was obtained by excluding the phases of acceleration and deceleration in isokinetic knee extensions and flexions at the highest velocity (180 deg-s -1) in which these phases are largest. In one patient, a smaller range, 45 to 80 °, had to be used owing to reduced range of voluntary movements. (EMG) The EMG signals were recorded with surface electrodes (Medicotest Type A-10-VS) b over the quadriceps and the
Eleetromyography
STRENGTH TRAINING IN STROKE PATIENTS, Engardt hamstring muscle groups. Skin preparation was performed by 95% ethanol washing and rubbing of the skin over the areas of the electrode placements. Two pairs of electrodes were placed as follows: quadriceps muscles: across the thigh, one third of the thigh length from the upper border of the patella and one fifth of the thigh circumference apart, and centered to a line between patella and the midpoint of ligamentum inguinale; hamstring muscles: along lines from tuber ossis ischii to the distal tendons of biceps femoris and semitendinosus at a distance from fossa poplitea 40% of the thigh length.~5 With these electrode positions, the total action of the knee extensor and flexor muscles is usually mirrored quite well. The EMG signals were amplified (1,000×) in miniature preamplifiers (Oxford HDX-82) a placed closely to the surface electrodes and fed to the standard EMG processing unit of the KIN-COM dynamometer system (KC/EMG)." In this, the EMG signals are integrated by rectification and bandpass filtering (20 to 1,000Hz) in a Paynter filter, followed by low-pass filtering (100Hz). 1637 The voltage signal thus obtained indicates the level of EMG activity. It was sampled at 100Hz using a 12 bit A/D converter. The integrated EMG was averaged for the same movement range as the one used to determine average torque.
10_ 98_ 7_
z5_
o4_
..J
i'
2 1 0
Body Weight Distribution on the Legs in Rising and in Sitting D o w n The vertical floor reaction forces under each foot during rising and sitting down were recorded with two force platforms. ~8 The patients sat in a standardized position on an adjustable chair with a back support. The seat height was set to the patient's knee height, determined as the distance from the lateral knee joint line to the floor, with the tibia perpendicular to the floor. The trunk was upright and the thighs supported for 3 of the femoral length. The patient placed one foot on each platform. The patients were instructed to stand up, to stand for 1 minute, and to sit down again. The instructions were "Please stand up, as you usually d o " and after 1 minute of standing, "Please sit down, as you usually do". A few pretest trials were allowed for the patients to become accustomed to the testing procedure. Strain gauge transducers fixed to each one of the platforms were used to record the vertical force under each foot during the test movement. The body weight distribution was estimated from the force load on the platforms during the test movements (ie, the time integral of the vertical forces under each foot). It was calculated from the area under the force curve from start to end of rising and sitting down, respectively (fig 1). The body weight distribution was defined as the ratio between the time integral of the vertical forces of the paretic and of the nonparetic leg. The means from three repeated tests of rising and of sitting down were calculated. The ratio of the impulses was expressed in percentage of body weight (%BW) supported by each leg. TM Gait The patients walked 30m under the following condition: (1) at their self-selected walking speed and (2) as fast as
421
0
I' 1
2
I 3
4
I 5
TIME (s) Fig 1--Areas under force-time curves (Ns) indicating the vertical forces of the nonparetic and the paretic leg, respectively, of a stroke patient when rising and when sitting down. Nonparetic leg, $~; paretic leg, i possible on an ordinary level floor outside the laboratory area. The gait was evaluated by measuring the mean walking speed and mean duration of the swing phase in percentage of the stride. One pair of pressure sensitive foot switches (Cefar) c was used to indicate the gait cycles of the paretic leg. One foot switch was taped to the heel of the shoe and the other to the sole of the shoe under the metatarsal head of the big toe of the paretic leg. The signals were recorded telemetrically (Medenik Biotelemetry System IC-600), d transmitted and fed to receivers, connected to ACamplifiers (Grass 7 P1F), e and printed on a polygraphic recorder. The collected data were processed in a computer-controlled digitizer system. For each gait cycle, the foot floor contact and the toe off of the paretic leg were marked on the stridecurve with a digitiser stylus for calculating the swing phase duration of the stride. 19
Training Device and Training Procedure Twice a week for 6 weeks, 10 patients trained the paretic leg exclusively with isokinetic maximal voluntary eccentric knee extensor actions and 10 trained exclusively with isokinetic maximal voluntary concentric knee extensor actions. A dynamic dynamometer (KIN-COM 125H) a controlled the movements. The patients sat as described under measurements of the torque (see previous section). After each active Arch Phys Med Rehabil Vol 76, May 1995
422
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
Table 2: Mean Torque (Nm) of Maximal Voluntary Eccentric and Concentric Knee Extensor Actions of the Paretic Leg at Three Different Constant Velocities Before and After the Training Period Eccentrically Trained (n = 10) Velocity Eccentric actions
Concentric actions
60 120 180 60 120 180
d e g ' s -j deg-s ~ d e g . s -~ d e g . s -~ deg-s J d e g . s -j
Before 106.3 113.8 114.4 62.4 47.2 35.3
_+ 12.9 ,+ 15.1 _+ 15.4 .+ 7.9 .+ 6.6 .+ 5.3
Concentrically Trained (n = 10)
After 136.6 142.5 142.3 79.1 59.1 44.2
p
+_ 15.8 .+ 17.8 ,+ 19.0 ,+ 9.5 ,+ 7.5 -4- 6.6
Before
<.01 <.01 <.01 <.05 <.05 <.05
109.4 112.8 112.4 61.8 44.1 29.8
After
,+ 11.0 _ 10.5 _+ 10.3 .+ 7.6 .+ 5.8 .+ 4.2
124.2 130.4 130.1 73.9 57.0 43.2
,+ ,+ .+ .+ .+ .+
p
11.2 11.0 10.3 8.0 6.1 4.7
<.01 <.05 <.05 <.05 <.01 <.01
Mean _+ SEM are given, p was calculated using Wilcoxon's signed rank test.
movement, the dynamometer moved the leg passively back to start position for the next active movement. The sessions started with 3 sets of 10 submaximal repetitions at 60deg. s -1, in the same mode as allotted for training, as a warm-up session. The range of motion of the knee joint was 10 to 100 ° except in a few patients, where the range was reduced and set to the patients' maximal capacity. Thereafter, the patients were asked to make a maximal contraction at 60deg. s -1, and a target curve was created. The training session consisted of repeated sets of 10 repetitions at angular velocities of 60, 120, 180, 120, 60, 120, 180deg. s -1, or more, with a 60-second rest interval after each selected velocity. To check that maximal effort was used the torque curves were superimposed on the first target curve on the monitor. The number of sets of repetitions increased similarly during the training period up to a maximum of 15 sets per training session. Each training session was followed by two 20-second maximal stretchings of the knee extensor muscles given by the physical therapist responsible for the training. Data Analyses
Differences within and between groups were tested for significance by the Wilcoxon's signed rank test and the Mann-Whitney U test. Nominal data were treated by the Chi-square exact test. 2° The level of significance chosen was 5%. RESULTS There were significant changes in mean maximum strength within both of the groups after training, as shown in table 2. In the eccentrically trained group, the mean changes were within the range of 30% to 25% for eccentric and concentric actions. In the concentrically trained group, the mean changes of torque were low (17% to 13%) in eccentric actions and high in concentric actions (25% to 57%). No statistically different changes between the two groups were found. Figure 2 shows the mean ___SEM values of the strength in the paretic leg relative to that of the nonparetic leg before and after the training period. Significant changes were found in the relative strength in the eccentrically trained group in both eccentric and concentric actions. The relative strength increased from 70%, 71% and 72% at 60, 120 and 180deg. s -1, respectively, before the training period to 89%, 88% and 86% after the training period in eccentric actions and from 62%, 54% and 50% to 76%, 62% and 60%, respectively, in concentric actions in the patients of the eccentrically Arch Phys Med Rehabil Vol 76, May 1995
trained group. Corresponding figures for the patients in the concentrically trained group were 78%, 76% and 78% before the training period and 85%, 84% and 86% after the training period in eccentric actions and 71%, 58% and 55% before the training period and 68%, 62% and 58% after the training period in concentric actions. The changes in the concentric group were not significant. Figure 3 shows the mean changes of the averaged agonist EMG activity (%) during voluntary knee extensor actions after training. In the eccentrically trained group, the mean changes were within the range of 33% to 24% for eccentric and concentric actions (p < .05). In the concentrically trained group, the mean changes were within the range of 28% to 23% in eccentric actions and 36% to 37% in concentric actions (p < .05). No statistically different changes between the two groups were found. Figure 4 shows the mean changes of the averaged antagonist EMG activity (%) during voluntary knee extensor actions after training. In eccentric actions, there was no change or a decrease of antagonist EMG activity in both the eccentrically and the concentrically trained group. However, in concentric actions, the antagonist EMG activity increased with increasing velocities in the concentrically trained group, whereas there was no change or a decrease in the eccentrically trained group. The mean changes of the antagonist EMG activity were significantly different in the concentrically trained group at the knee angular velocity of 120 and 180deg • s- 1compared with those of the eccentrically trained group. A o~
Eccentrically Trained ]
j
j
j
.
~
B
j
80
80
60
60
~o 40
40
~
20
o
.
20 0
-180
-120
-60
Eccentric
Concentrically Trained
60
120 Concentric
Velocity (deg/s)
180
0
.Before [ ] After
-le0
-120
-60
Eccentric
60
120 180 Concentric
Velocity (deg/s)
Fig 2 - - T o r q u e of maximal voluntary eccentric and concentric knee extensor actions in the paretic leg relative to that of the nonparetic leg at three different constant angular velocities in (A) the eccentrically (n = 10) and (B) the concentrically trained group (n = 10) before and after 6 weeks of isokinetic training. Values are mean +__SEM. P was calculated using the Wilcoxon's signed rank test. *p < .05, **p < .01.
STRENGTH TRAINING IN STROKE PATIENTS, Engardt Voluntary knee extensor action
423
Rising
Sitting down
A
o~
60
---
_>, ._> ~
ECC. tr.
m
60
40
I$
ro
50
-:-- ~
40
:~ -~
30
.-~ ~ c-,
20
o~,^.^
W to
~
20
o or--
t-0
0
I
m
-180-120-60
|
60
Eccentric
120
180
Concentric
Velocity (deg/s) Fig 3 - - C h a n g e s of agonist EMG activity during maximal voluntary eccentric and concentric knee extensor actions at three different constant angular velocities in the eccentrically (n = 10) and the concentrically (n = 10) trained group after 6 weeks of isokinetic training. Values are mean _+ SEM. P was calculated using the Mann-Whitney U test. The changes in the two groups were not significantly different.
Figure 5 shows the body weight distributed on the paretic leg (%BW) in rising and in sitting down before and after the training period. A significant improvement towards symmetrical body weight distribution on the legs in rising was noted only within the eccentrically (p < .01) trained group. Voluntary knee extensor action
A
v
t
g" ._>
so
w
0
40
IJ.I
"F
O c~
[]
ECC. tr.
•
CON. tr.
7
20
e-
~
0
o
" t-O
10
nq o
r"
B
-20
|
-180 -120
|
-60
|
60
120
180
Eccentric Concentric Velocity (deg/s) Fig 4 - - C h a n g e s of antagonist EMG activity during maximal voluntary eccentric and concentric knee extensor actions at three different constant angular velocities in the eccentrically (n = 10) and the concentrically (n = 10) trained group after 6 weeks of isokinetic training. Values are mean _+ SEM. P was calculated using the Mann-Whitney U test. The changes in the two groups were not significantly different except in concentric actions at 120 and 180deg. s -1. *p < .05.
0 Ecc.tr.
Con.tr.
Ecc.tr.
Con.tr.
Fig 5 - - B W distribution on the paretic leg in percent of the total body weight in rising and sitting down in the eccentrically (n = 10) and the concentrically trained group (n = 10) before and after 6 weeks of training. Values are mean _+ SD. P was calculated using the Mann-Whitney U test. *p < .05.
The change in the eccentrically trained group differed significantly (p < .05) from that of the concentrically trained group. No improvements of weight distribution on the legs were observed in sitting down in any of the two training groups. Significant changes were noted in walking speed and in duration of swing phase within the concentrically trained group (p < .05). The patients walked faster but loaded their paretic leg for a shorter period of the walking cycle after training. However, the changes in the two groups were not significantly different, as shown in table 3. DISCUSSION The present study shows that eccentric as well as concentric training gave a considerable increase of knee extensor strength after six weeks training. The increase in strength was related to enhanced activation of the agonists as judged from the mean changes of the averaged EMG activity. The changes in the averaged EMG agonist activity were proportional to the changes of torque in eccentric as well as in concentric actions in both groups. This indicates that the strength increase was caused by neural factors. The intensive loading through the isokinetic training might activate previously dormant fibers. 2~ The increase in strength production might be caused by an increased ability to raise motoneuron activation during voluntary effort. 22 This is in accordance with results after training in healthy subjectsY 27 As shown in the present study, there were significant improvements in the relative strength of the paretic leg in eccentric as well as concentric actions after eccentric training but not after concentric training. In an earlier study of patients with spastic paresis, it was shown that there may be considerable restraint to maximal voluntary concentric movements because the spastic reflexes in the antagonists may be facilitated by strong voluntary effort. ~ The repeated forceful voluntary movements of concentric training will Arch Phys Med Rehabil Vol 76, May 1995
424
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
Table 3: Walking Speed (m/s) and Swing Phase Duration (% Gait Cycle) of the Paretic Leg at Self-Selected (A) and Fast Speed (B) in the Eccentrically and the Concentrically Trained Group Before and After Training Eccentrically Trained (n = 10) Before Walking A (m/s) Walking B (m/s) Swing A (% stride) Swing B (% stride)
.81 1.0 43.9 45.1
± .18 ± .25 _+ 6.4 ± 5.9
After .84 1.0 43.9 46.6
_+ .24 _+ .26 + 4.6 ± 4.4
Concentrically Trained (n = 10) p ns ns ns ns
Before .65 .81 42.1 45.1
+_ .2 +_ .29 ± 5.6 ± 6.1
After .73 .91 45.5 47.1
± .25 ± .30 _+ 4.3 ± 6.2
p <.05 <.05 <.05 ns
Note: Changes between the two groups were not statistically different. Mean + SD are given, p was calculated using Wilcoxon's signed rank test and the Mann-Whitney U test.
imply repeated relatively strong reflex activations of the antagonists that may explain the increased antagonist EMG activity after concentric training. An increase in agonist activation caused by training will thus in part be out balanced by increased antagonist reflex activation, which also appears to be an effect of concentric training. In contrast, the antagonists are not activated by stretch reflexes in eccentric training. If stretch reflexes are activated, they are activated in the agonist that is lengthening. Thus, the stretch reflexes will act in synergy with the voluntary effort. These differences may be the reason why eccentric training had a better effect on the voluntary strength. The difference between the effects of eccentric and concentric training appears to be of importance for motor functions of daily life. Thus, the training had different effects on the capacity to load weight symmetrically on the legs during rising up from a sitting posture. After the eccentric training but not after the concentric training, there were significant improvements towards symmetrical body weight distribution in this motor task. It may be a consequence of the larger increase of the strength of the paretic legs relative to the nonparetic after training. It appears likely that the lesser the difference is between the strength of the two lower extremities, the lesser the risk should be of a continued learned disuse of the paretic leg. The action of the quadriceps muscles in rising from a sitting position is a concentric one. The fact that this function was improved after eccentric training implies that the eccentric training had effects not only on the eccentric function trained but also gave a strength increase that could be utilized in concentric actions. No improvements in BW distribution in sitting down were noted after training. However, the distribution of body weight in the patients of the present study was almost equivalent to that of old healthy subjects presented in an earlier study.~8 Thus, an already close to symmetrical relearned motor program of sitting down from standing was not influenced by the improvement of the knee extension torque on average 32 months after stroke. Despite significant improvements in mean knee extension torque, there were no significant changes in the gait parameters except for a significant improvement in walking speed and a prolonged swing phase duration in the concentrically trained group. Mean gait velocity in old healthy men is 1.18m/s (age 67 to 73 years) and 1.45m/s (age 60 to 65 years). 28 As the patients in the eccentrically trained group walked on average with 0.81 at self-selected and 1.0m/s at fastest speeds possible already before training, there might Arch Phys Med Rehabil Vol 76, May 1995
not have been much room or motivation for an increase of the gait velocity. In the concentrically trained group, where the gait velocity was low, the training, however, had an effect. The results of the present study give support for the concept that eccentric training may be more suitable for stroke patients than concentric training. Eccentric training to improve motor functions after stroke was suggested already by Brunnstrtm. 29 However, manual resistance cannot give the same specificity in training. With the isokinetic dynamometer used, movements can be controlled at well-defined movement velocities and maximal contractile tension within a large movement range. Thus, there is a pronounced difference between dynamic training when controlled manually and controlled by a strong active dynamometer. The latter has sufficient capacity to overpower or resist maximal voluntary actions. This implies that eccentric and concentric actions can both be trained at the maximal levels of contractile tension. References 1. Knutsson E, M~rtensson A. Dynamic motor capacity in spastic paresis and its relation to prime mover dysfunction, spastic reflexes and antagonist co-activation. Scand J Rehab Med 1980; 12:93-106. 2. Freund HJ. The pathophysiology of central paresis. In: Struppler A, Weindl A, editors. Electromyography and evoked potentials. Berlin: Springer-Verlag, 1985; 19-20. 3. Langton-Hewer R. Rehabilitation after stroke. Quart J Med 279:65974, 1990. 4. Bobath B. Adult hemiplegia: evaluation and treatment. 2nd ed. London: William Heinemann Medical Books Ltd, 1978:58-63. 5. Bohannon RW. Muscle strength in patients with brain lesions: measurements and implication. In: Harms-Ringdahl K, editor. Muscle strength. New York: Churchill Livingstone, 1993; 187-225. 6. Baker P, Rawicki B. The effect of movement, gait and resistance to stretch of spastic muscle following an isokinetic strengthening program. Proceedings of the Xlth International Congress of World Confederation for Physical Therapy and Chartered Society of Physiotherapy. London: 1991:1009-1111. 7. Knutsson E, M~rtensson A, Gransberg L. The effects of concentric and eccentric training in spastic paresis. Scand J Rehab Med 1992;24(27 Suppl):31-2. 8. Knutsson E. Concentric and eccentric muscle work in spastic paresis. Scand J Rehab Med 1992;24(27 Suppl):16-7. 9. Fugl-Meyer AR, J~iask6 L, Leyman I, Olsson S, Steglind S. The poststroke hemiplegic patient. I. a method for evaluation of physical performance. Scand J Rehab Med 1975;7:13-31. 10. Duncan PW, Propst M, Nelson SG. Reliability of the Fugl-Meyer assessment of sensorimotor recovery following cerebrovascular accident. Phys Ther 1983;63:1606-10. 11. Can" JH, Shepherd RB, Nordholm L, Lynne, D. Investigation of a new motor assessment scale for stroke patients. Phys Ther 1985; 65:175-80. 12. Poole JL, Whitney SL. Motor assessment scale for stroke patients:
STRENGTH TRAINING IN STROKE PATIENTS, Engardt
13. 14. 15. 16. 17.
18. 19.
20.
21. 22.
concurrent validity and interrater reliability. Arch Phys Med Rehabil 1988; 69:195 -7. Ashworth B. Preliminary trial of carisoprodol in multiple sclerosis. Practitioner 1964; 192:540-2. Engardt M, Knutsson E. Dynamic thigh muscle strength after auditory feedback training of body weight distribution in stroke patients. Physiother Theory Practise 1994; 10:103-12. Knutsson E, Richards C. Different types of disturbed motor control in gait of hemiparetic patients. Brain 1979; 102:405-30. Gottlieb GL, Agarwal GC. Filtering of electromyographic signals. Amer J Phys Med 1970;49:142. Gransberg L, Knutsson E, Litton JE. A computer programmed system for the analysis of active and passive isokinetic movements. In: IEEE frontiers of engineering in health care. Proceedings, Institute of Electrical and Electronics Engineers/Medicine and Biology Society, Washington, DC. 1980;292-295. Engardt M, Olsson E. Body-weight bearing while rising and sitting down in patients with stroke. Scand J Rehab Med 1992;24:67-74. Myrenberg M, Lundeberg T, Od6en I. Effect of traction treatment on EMG activity during walking in erecti spinae in patients with low back pain. The Pain Clinic 1990;3:67-73. Mehta CR, Patel NR, Tsiatis AA. Exact significance testing to establish treatment equivalence for ordered categorical data. Biometrics 1984;40:819-25. Perrine JJ. Isokinetic exercise and the mechanical potentials of muscle. JOHPER 1968;5:40-5. Sale DG, MacDougall JD, Upton ARM, McComas AJ. Effect of
23. 24.
25. 26. 27. 28. 29.
425
strength training upon motoneuron excitabilituy in man. Med Sci Sports Exerc 1983; 15:57-62. Barnes W. The relationship of motor-unit activation to isokinetic muscular contraction at different contractile velocities. Phys Ther 1980;60:1152-8. Milner-Brown HS, Stein RB, Lee RG. Synchronization of human motor units: possible roles of exercise and supraspinal reflexes. Electroencephalography Clin Neurophysiol 1975;38:245-54. Rutherford OM, Jones DA. The role of learning and coordination in strength training. Eur J Appl Physiol 1986;55:100-5. HSkkinen K, Komi PV. Electromyographic changes during strength training and detraining. Med Sci Sports Exerc 1983; 15:455-60. Moritani T, De Vries HA. Reexamination of the relationship between the surface integrated electromyogram (IEMG) and force of isometric contraction. Am J Phys Med 1977;57:263-77. Murray MP, Kory R, Clarkson BH. Walking patterns in healthy old men. J Gerontol 1969;24:169-178. Brunnstr~m S. Movement therapy in hemiplegia. New York: Harper & Row, 1970:69.
Suppliers a. Chattecx Corp,, 101 Memorial Drive, PO Box 42887, Chattanooga, TN 37405. b. Medicotest Type A-10-VS. c. CEFAR medical products, S-222 41 Lund, Sweden. d. Medenik AB, S-740 60 t3rbyhus, Sweden. e. Grass Instruments Corporation, Quincy, MA. (in Sweden: Somedic AB, tJstermalmsgatan 27, S-123 42 Farsta, Sweden).
Arch Phys Med Rehabil Vol 76, May 1995