Improvement in hand function and sensation in chronic stroke patients following electrical stimulation exercises. A pilot study

Paul Taylor, Jane Burridge, Sean Hagan, Paul Chappell*, Ian Swain.

Department of Medical Physics and Biomedical Engineering, Salisbury District Hospital, Salisbury, Wiltshire, SP2 8BJ

* Dept of Electrical Engineering, University of Southampton, Southampton, S017 1BJ.


Objective: To measure the effect of electrical stimulation exercises of the hand on function, grip strength and sensibility.

Design: Retrospective analysis of clinical treatment.

Subjects: 20 hemiplegic patients who have suffered a stroke at least six months prior to commencement of treatment.

Setting: Outpatient bassed, exercises perfomed at home

Intervention: Electrical stimulation of finger, thumb and wrist extensors and flexors.

Main outcome measures: The Jebsen-Taylor hand function test, palmer, lateral, and pinch grip strength, and static two point discrimination.

Results: 15 subjects improved their Jebsen-Taylor test score while 3 reduced their score and 2 were unchanged. The 16 subjects who were unable to complete the full test at the start of the trial, were able to complete an average of an additional 3.4 tasks, an increase of 24.7% (p<0.001) There was a significant increase in key grip of 38 % (p < 0.01). The mean two point discrimination score increased from 1.80 to 2.44 (p<0.02). There was a significant difference between affected and non affected hands at the start of the treatment but no statistical difference at the second static two point discrimination score.

Conclusion: The results suggest that there was a training effect from using electrical stimulation. It is not clear to what extent this benefited activities of daily living. Further work is needed to verify these results in a more rigorously controlled trial.


Approximately 100,000 first ever strokes occur in Britain each year.1 Approximately half of all acute stroke patients starting rehabilitation will have a marked impairment of function of one arm and only about 14 % of these will regain useful function.2 Electrical stimulation of the finger, thumb and wrist extensors has bean shown to improve the passive range of wrists extension in 16 subjects3 following 4 weeks of electrical stimulation exercise for three thirty minute periods a day. While a trend was shown to increased voluntary wrist extension torque and a reduction in spastisity (measured on a four point scale), these changes were not statistically significant. Sensation was measured using two point discrimination and was not changed by the electrical stimulation. The effect on passive wrist extension was shown in a second study4 after two weeks of wrist extension exercises for four periods of thirty minutes a day. Eleven subjects took part in an A-B-A style trial to assess the longer term benefits of the therapy. The improvement in wrist extension was lost after just one hour after the final exercise session and unchanged two weeks later. Improvement in voluntary wrist extension and torque have been reported by a third suddy5 using electrical stimulation as a feed back parameter in a biofeed back device. The subject's arm was held in a rig which measured wrist extension. When voluntary wrist extension exceeded a set threshold, a stimulator was triggered to produce full wrist extension. In this way reinforcement was given to the production of the desired movement.

Non of the aforementioned papers have quantified any functional benefit to the recipients of the therapy. Only on study has looked at this aspect6 which used the Fugl-Meyer post stroke motor recovery test and Jebsen-Taylor hand function test. The study compared three groups who received proprioceptive neuromuscular facilitation therapy, bias balance electrical stimulation therapy or EMG initiated electrical stimulation of the wrist extensors and other muscle groups with a control group. Bias balance therapy consisted of five thirty minute periods of low intensity stimulation to wrist extensors. The intention of this stimulation was not to elicit a movement but to cause a relaxation of flexor spastisity. EMG triggered stimulation was triggered by low level activity in the paralysed muscles to be stimulated and carried out in three sessions per week under therapist supervision. All groups received three months of therapy. Significant improvements in FM score was seen in all three treatments with the greatest gains being seen in EMG triggered stimulator group followed by the bias balance stimulation group. The controls showed no improvement. Those subjects in the treatment groups who could complete a Jebsen test also showed improvement.

On study has shown changes in perception following electrical stimulation7. Two subjects received "contingency stimulation " training for two period of 30 days preceded and separated by two thirty day base line periods. Electrical simulation of the forearm was triggered by movement of the non affected arm, to remind the subject of their affected limb each time the other was used. Both subjects showed statistical significant improvements in the Rivermead Perceptual Assessment Battery in the treatment phase which was maintained in the second base line period.

A seventy year old right CVA subject, seven years post CVA was referred to our department for electrical stimulation exercises to strengthen and improve the fatigue resistance of his wrist extensors. The wrist extensors were exercised using a Raymar Pocket Orthotron neuromuscular stimulator using 300 micro second pulses at twenty pulses a second, for two periods of twenty minutes a day. The stimulator gave a duty cycle of eight seconds on and ten seconds off. After one month the subject reported a significant improvement in not only wrist function but also finger extension. The subject reported improved hand function such as doing up buttons, putting on his reading glasses and tying shoe laces. He also believed his ability to feel had been improved. These reports, while subjective, convinced us that the use of electrical stimulation exercises to improve CVA hand function should be investigated further.


Twenty subjects who were undergoing treatment within the department of Medical Physics and Biomedical Engineering were followed in this study. All subjects had had a stroke, 8 of whom had right sided hemiplegia and 12 left sided hemiplegia. 14 subjects were mail while 6 were female. The mean age at the commencement of treatment was 49.2 Years (sd 15.6 years) and the mean time since stroke was 5.9 years (sd 7.1 years). All subjects had restricted hand and arm function but were able to bring their hand to their mouth.

Materials and Method

The subjects received electrical stimulation exercises of the hand and wrist muscles. Wrist and finger extensors were stimulated reciprocally with lumbricals, finger flexors or triceps. The exact pattern of exercise varied from individual to individual being based on the clinical judgement of the clinician directing the treatment. Subjects used the Microstim 2 or Odstock Four Channel stimulators, both designed and manufactured within the department. Both devices produced up to 100 mA with a pulse width of 300 m s and pulse repetition rate of 40 Hz. The contraction length was 8 s with a 10 s rest period. The output was rammed over 2 s at the beginning and end of each contraction for comfort and also to attempt to prevent stretch reflexes in the antagonist muscles in response to sudden movements. Pals Plus self adhesive skin surface electrodes were used. Subjects were asked to exercised twice daily at home for up to one hour per session.



Three assessments were made. The Jebsen - Taylor Hand Function Test8 consists of seven individual tasks and is performed by each hand. Each task is timed and if not completed within the maximum allowed time of 80 s, the number parts of the task completed recorded (maximum 30). The tasks were:

Writing. A standardised text is copied.

Turning 5 cards (75mm x 100mm).

Picking up and placing in a tin (180mm x 100mm dia.) two bottle tops, two USA pennies and two paper clips.

Stacking four draughts pieces on a board (5mm x 20mm dia.).

Simulated feeding. Using a tea spoon, five kidney beans are lifted and dropped in to a tin.

Lifting five empty 1 lb. empty baked bean tins (115mm x 75mm dia.) on to a board, open end down.

Lifting five full tins. As above but with unopened tins.

No set method was proscribed for each test. Rather the subject was free to choose what ever method they found most efficient. In this way it was hoped that the test would reflect more closely tasks in every day life. It was found that most subjects were unable to perform the writing test which in any case may be considered meaningless when it is the non dominant hand that is being tested, so this was omitted from the analysis of the results.

Sensation was tested using static two point discrimination 9. The hand was divided up into 24 areas, four palm areas and four areas per digit (fig.1). The subject was blindfolded and asked to say whether they perceived one point or two when the hand was touched with the probe. The hand was scored as follows:

0 = no sensation,

1= sensation but no discrimination or ambiguous answers given,

2 = 10mm discrimination

3 = 6mm discrimination

4 = 4mm discrimination

5 = 2mm discrimination

Each area was tested using the 10mm probes first and the probe size reduced until no discrimination was possible. The subject was also tested with one point randomly throughout the measurement. The mean two point discrimination (the sum of the scores divided by 24) was recorded for each hand.

Grip strength was recorded using an MIE Grip Strength Analyser. Power, pinch and key grip were tested for each hand. Subjects were allowed three attempts and the greatest measurement recorded.

Subjects were tested prior to exercise and again between two and three months after exercise. All measurements were made before any electrical stimulation had been used on that day. Results are given below. It was not possible to record all measurements on all subjects. Subject numbers have been given in each case.


Paired t tests were used to compare groups of data. Significance was assumed if p<0.05. Normal distribution was demonstrated by examining mean, median and modes and by using skewness tests. Paired T test were performed on the static two point discrimination results from each hand, comparing the 24 areas of the hand before and after treatment. A probability of p < 0.05 was be taken as signifying a statistically significant change in any one subject.



Jebsen test.

Overall, 15 subjects had improved scores, 3 deteriorated and 2 had no change. Four subjects completed all the tasks before and after, reducing their mean time per task by 37.9% (graph 1). The 16 subjects unable to complete all the tasks before treatment, were able to complete on average 3.4 additional tasks by the second measurement, an increase of 24.7% (paired T test p < 0.001) (graph 2). There was no significant change in the unaffected hand.





































Table 1. Jebsen test results

M0 - measurements taken at the first assessment

M3 - measurements taken at two - three months

Grip strength.

There was a significant increases in key grip (n=9, 38.3% increase, p<0.01). There was also an increase in the power grip of the unaffected hand. (n=13, 15.1% increase, p < 0.01.)


Power grip (N)

Pinch grip (N)

Lateral grip (N)



3 months


3 months


3 months




































Table 2 Grip strengths, affected hand.



Power grip (N)

Pinch grip (N)

Lateral grip (N)



3 months


3 months


3 months




































Table 3 Grip strength unaffected hand


Static Two point discrimination.

The mean static two point discrimination score changed from 1.80 to 2.44 between the assessments (p < 0.02). 7 subjects increased their score, 4 were unchanged while 0 reduced their score. No significant change was found in the unaffected hand, changing from 2.91 before electrical stimulation to 2.78 after treatment. There was a significant difference between the affected and non affected hands at the start of the study but no statistical difference at the second measurement. Graph 3 summarises the results from the 11 subjects showing mean scores for the affected and unaffected hand before and after electrical stimulation exercises.






















Table 4 Static 2 point discrimination (n=11)

M0 - measurements taken at the first assessment, M3 - measurements taken at two - three months


While overall statistically significant improvements were recorded, as there was no control group it can not be ruled out that these changes were not due to the natural progression of the subjects condition. Nor were the assessments performed blind. It is not clear if these improvements were carried over into every day life. Some subjects reported improved function, for example one subject was able to do up his trouser buttons and belt using both hands, another was able to use a knife and fork, another manipulate a pepper grinder, and another was able to write all the families Christmas cards where previously they had only been able to do one or two. However some subjects received no improvement and none had a return to normal function. No relationship can be found between the performance in the tests and the time since CVA, age or side of hemiplegia. There is also no indication that subjects with an initial very poor hand function improved their test scores any differently to subjects with initially reasonable hand function. Backer et al.3 did not record changes in hand sensibility. This may be due to a shorter treatment period in their study or diferences in the way two point discrimination was caried out.

It is not clear what the mechanisms behind these changes may be. The fact that a change in two point discrimination was recorded suggests that neural-plastic changes in the brain may have occurred and this may be due to increased activity in the hand while performing the exercises. It is possible that repeated stimulation of finger, thumb and wrist extensors may cause habituation of reciprocal inhibition in the antagonist muscles leading to a reduction in spastisity 10. This may lead to improved control of the hand. It is also possible that there has been a reduction in neglect syndrome. While some increase in grip strength was recorded it is important to remember that the typical stroke subjects has significantly more difficulty in releasing his grip than in making a grip. Further work is required to validate these results in a more rigorously controlled trial and with a greater number of subjects, to establish which patient group can best benefit from this treatment and determine if the benefits are translated into improved function in daily living.

Address: Department of Medical Physics and Biomedical Engineering, Salisbury District Hospital, Salisbury, Wiltshire, SP2 8BJ, UK. Tel. 01722 336262 Ex. 4065


We would like. to thank the department of Health, Medical Devices Agency, the charity Inspire for there support of this work and the Wessex Rehabilitation Association for allowing us to inhabit their building.


1. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerbrovascular disease in the community. The Oxford Community Stroke Project 1981-86. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerabral and subarachnoid haemorrhage. J. Neurol Neurosurg Psychiatry 1990 53:16-22

2. Wade DT, Langton-Hawer R, Wood VA, Skilbeck CE and Ismail HM. The Hemplegic arm after stroke: measurement and recovery. J Neurol Neurosurg Psychiatry 1983; 46: 521-524

3. Baker LL, Yeh C, Wilson D, Waters RL. Electrical stimulation of wrist and fingers for hemiplegic patients. Physical Therapy 1979; 59 (12): 1495 - 1499.

4. Pandyan AD, Granat MH. Effects of electrical stimulation on flexion contractures in hemiplegics. Clin Rehab 1997; 11: 123 - 130

5. Bowman BR, Baker LL, Waters RL Position feedback and electrical stimulation: an automated treatment for hemiplegic wrist. Arch Phys Med Rehabil Vol.60 Nov 1979 497 - 501.

6. Kraft GH, Fitts SS and Hammond, MC. Techniques to improve function of the arm and hand in chronic hemiplegia. Arch. Phys. Med. Rehabil. 1992; 73: 220-227.

7. Prada G, Tallis R. Treatment of the neglect syndrome in stroke patients using a contingency stimulator. Clin Rehabil 1995; 9: 304 - 313

8. Jebsen RH, Taylor N et al. An objective and standardised test of hand function. Arch. Phys. Med.and Rehab. 1969; June: 311-319

9. Lee Dellon, A. The sensational contributions of Erik Moberg. J. Hand Surg. 1990; 15B: 14-24.

10. Burridge J, Taylor P, Hagan SA, Wood DE, Swain ID. The effect on the spasticity of the quadriceps muscles of stimulation of the Common Peroneal nerve of chronic hemiplegic subjects during walking. Physiotherapy vol. 83, no 2 1997